Hacker News new | past | comments | ask | show | jobs | submit login

"LED's are nearing theoretical max efficiency, somewhere near 220lm/watt for pure white light."

Cree has already hit 300+ lumens per watt at room temperature and 1W power throughput, and are showing no signs of slowing. The problem with LEDs is not wave issues, but the Auger Effect where electrons recombine with the LED junction material instead of going through the electron hole.

However, CRI is still meh. For gem work, you want sunlight or incandescent, nothing else is acceptable because you need a true blackbody radiator. LEDs work fine for isolating and orienting stars, adularescence, schiller, and sheens, but are horrible for tenebrescent gems, some dichroic gems, and such.

Also, some LED-covered wavelengths are still utterly useless when compared to other sources. UVB LEDs at their best are only about 2% efficient. You might as well stick with a raw phosphor-less fluorescent tube with glass filters for UVC and UVA on it at that point, you'd still push roughly 30% efficiency.

Red LEDs are also still rather meh on the efficiency scale.




Yeah, the main problem with LEDs is that you only get max efficiency if you can find something with the right bandgap. Right now that means deep blue.

220 w/lm is theoretical max for a smooth spectrum of white light. If you get to choose a spectrum you can get much higher using green centered at 555nm. If any "white" light advertises close or above 220 lm/watt the CRI suffers. It means they're using light that activates the human retina better but isn't actually a pure white.

Lumens are calculated based on human eye response https://en.m.wikipedia.org/wiki/Luminosity_function

Looked at this way blue lights are pretty terrible. But if you look at mw of photons vs energy input deep blue LED's are close to 80% efficient.




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: