There was a time when LED light bulbs were a premium product that commanded a premium price, mainly because of limited supply and the usual marketing tricks. But now is not that time, since you can pick up an LED bulb for a buck or two at pretty much any store. So why in the world would you go to the effort to make your own light bulb?
For [DiodeGoneWild], the answer is simple: it’s all about staying in rhythm. Circadian rhythm, that is. We all know how light toward the blue end of the spectrum is bad for our sleep cycle, since it convinces our lizard brain that dawn is at hand. But even if you pick an LED bulb with a warm, or reddish, color temperature, there’s still a lot of UV light being emitted thanks to the phosphor LEDs that are typically used in them.
[DiodeGoneWild]’s first attempt at a design, in the first video below, mostly avoids phosphor LEDs in favor of a mix of yellow, red, and yellow-green LEDs to get a warmer spectrum. He used the housing and base from an expired bulb to enclose his custom circular PCB, the fabrication of which using a hand drill as a lathe and a Dremel to machine concentric tracks in the cladding was a real treat. So was the power supply, for that matter — a dropping capacitor followed by a bridge rectifier and a filtering cap. We like the discharge resistors across the caps and the fusible resistor on the mains side — it’s nice to see safety factored in from the start. And what’s not to like about using a DVD as a makeshift spectroscope?
We see that [DiodeGoneWild] has just dropped a second design, this time in a much smaller bulb and with relatively more phosphor LEDs.
Most human eyes have 3 cones, for R, G, and B, and the blue cone has a peak response around 437 nm. (Some people have a 4th cone sensitive in the UV.)
The circadian rhythm is sensed by ganglia in the retina itself. This is separate from the B cone signals used in perception, is completely unconscious, and is at a different wavelength (470 nm) than the B cone.
Additionally, the cornea and lens become progressively opaque to the blue light as we age. This paper (https://bjo.bmj.com/content/92/11/1439) states that “A 10-year-old child has circadian photoreception 10-fold greater than a 95-year-old”.
If you want to experiment with this easily and cheaply, get red laser protection goggles from eBay and wear them an hour to 90 minutes before you go to bed.
Works a treat, and can be used as a starting point for life-hacking/experimentation.
As the videos note, most LEDs emit substantial blue light that can confuse circadian reception, even the “warm red” types.
P.S. – Love the build in this, especially making a round PCB and using an old CD disk for spectral analysis of the various lights.
True hacker material.
Red or Blue glasses colour?
Red
@PWalsh, informative comment – thank you.
So do you need the red ones or the blue ones, I feel I should know this but not totally sure.
You want the ones that look red to your eyes. :) I use the Uvex ultraviolet blocking ones that look orange. Same concept, blocks nearly all uv and blue light
Genius. I might try that with my laser goggles
It is actually well studied and known:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4734149/
Blue light suppresses melatonin in wild animals:
https://phys.org/news/2022-09-dark-side-suppression-melatonin-blue.html
Night-time blue light pollution increases across Europe:
https://www.science.org/doi/10.1126/sciadv.abl6891
https://www.nature.com/articles/s41598-018-36791-5
In 2002, it was established that a novel, third type photoreceptor regulates a large variety of biological and behavioural processes1,2, including melatonin3,4,5,6 and cortisol5,6,7,8 secretion, alertness…
So unless you cite me an article that disproves 20 years of research, I’m gonna side with the article author on this one.
Does being judgemental help ?
What is wrong with using an efficient white LED and blocking out the frequencies you do not want with a filter ? (e.g. an additive like goyenchem-bl475 which absorbs up to 475nm)
Would that use less power or more power ?
I’m thinking about cheap theatrical gels (e.g. https://en.wikipedia.org/wiki/Color_gel ) or expensive optical filter ( https://en.wikipedia.org/wiki/Optical_filter )
The manufacturers themselves might be able to help with this, specially about the uv, by using additives on the lenses, or changing the material itself, like using polycarbonate.
Filtering out wavelengths will reduce optical output, so, it will require more power input to compensate. Unless the dye has fluorescent characteristics.
They’ve determined quite a few years ago that blue light has pretty much ZERO impact on sleep cycles, etc. and that instead it’s exposure to brighter light from our devices that interferes. You don’t have to believe me, just listen to the people of the Wills Eye Institute(established almost 200 years ago and THE place to go for legit info about the eyes)…………….
https://www.youtube.com/watch?v=NkJY9bgLyBE
It’s actually the blue spectrum that mimics a daytime signal and does disrupt the sleep patterns. In senior assisted living facilities you can reduce falls by up to 43% when giving them a true daytime signal of 490nm Spectrum as it permeates through the yellow lens of their eye up to 50% more than legacy led 450nm. This then helps them take less naps, sleep through the night and that was when they were having their falls
They have discovered ganglion (IPRGC) cells that are non visual and prefer the 490nm that is the peak sensitivity in humans, plants and animals. The legacy led bulbs have 450nm and that is the bad blue in the sky, amber is more about night time. You need 490nm for true daytime signal – 150 EML minimum and then only 11 EML about 2-3 hours before bed. This is a daytime, nighttime, reset of your circadian clock