Making LED bulbs safer: Health and LEDs

I am thinking about the health aspects of home led lighting and whether it is possible to filter out the shorter wavelength light. I am not sure what manufacturers do in order to produce their 'warm light' bulbs. They claim a lower colour temperature but how are they achieving that?

Welcome to PF.

Are you thinking about eye strain from too much blue from some LED light fixtures? What health effects specifically are you concerned about? Can you provide some links to the reading you've been doing about this? Thanks.

hi and thanks for your reply. My understanding is that the shorter wavelengths (350-450nm.) destroy melatonin which leads to poor sleep and can result in macular degeneration. The 'soft white' bulbs claim a colour temperature of 1800- 2000 °K. but I don't know if that is even correct - or how it is achieved.

I am thinking that the semiconductor producing the light continues to produce between 350 - 700 nm. so my difficulty is understanding how only the longer wavelengths emanate from the so called 'warm white' bulbs.

I can't believe that the semiconductor is changed, nor can I believe that the wavelength of light it emanates can be changed so what are the manufacturers doing to achieve this yellow light. Using some kind of phosphor I guess - but I don't understand the process.
cheers

renault said:

so my difficulty is understanding how only the longer wavelengths emanate from the so called 'warm white' bulbs.

Something you need to realise about our perception of colour is that when we see a reddish or bluish colours (such as the sunset or a blue sky) the spectrum that we are seeing is actually broad band and contains all visible wavelengths with just a 'slope' over the spectral band. Blues and reds are seldom removed from any common light source - except for monochromatic sources and suitably driven TV displays.

Some while ago, I bought myself a small (pen-top size) spectroscope and I've been looking at the spectra of many light sources. The light from a good (i.e. convincing) domestic LED has a much more uniform spectrum than the sunlight we see because it's not affected buy the atmospheric absorption notches in the visible range. Same thing goes for a good tungsten filament. otoh, CFLs are totally disgusting and the spectrum looks unbelievably ridiculous.

sophiecentaur said:

The light from a good (i.e. convincing) domestic LED has a much more uniform spectrum than the sunlight we see because it's not affected buy the atmospheric absorption notches in the visible range.

Could you post the spectra from a couple of your LED light bulbs in your home, along with their ratings (power, "warmth", etc.)? That would be interesting to see...

yes indeed that would be interesting to see. I still don't have my question answered though - that is; the so called 'Edison' old style new LED bulbs clearly appear to show a yellowish light - and they claim a colour temperature of 1800°K. So how do you think they are doing this?

berkeman said:

Could you post the spectra from a couple of your LED light bulbs in your home, along with their ratings (power, "warmth", etc.)? That would be interesting to see...

The device is very small and works well when you put your eye right next to it - very impressive in fact. But I never go round to fixing it securely enough to take a photo image.
First of all, I thought there were 'imperfections' in the way it shows the spectrum of the Sun (was it the glass in the window?? etc.) but my new lcds (warm, 3500k) show no such stripes / bands so I have to conclude that I'm actually looking through a good instrument.
I guess I should grasp the nettle and clamp my iPhone to it so that I can do a rough measurement of actual levels.
I seriously recommend you guys should all spend £30~ (eBay) and get one. Hours of fun!!

renault said:

the so called 'Edison' old style new LED bulbs clearly appear to show a yellowish light - and they claim a colour temperature of 1800°K. So how do you think they are doing this?

Afaik, domestic lighting bulbs start with a UV source which passes through a mix of phosphors to produce any spectrum they want. If you're talking about the 'false filament' lamps then

If you look on the Wiki information about leds in general (plus loads of other sites) there is information about basic monochromatic leds and the history of development from the original red ones. There are systems with three colour sources which can be made to 'look' any colour you want but, unless the sources are broad band then the resulting light is not good as an illuminant; a continuous spectrum is needed for that. I'd imagine that those retro lamps are produced to look nice rather than as a good low temperature illuminant.

berkeman said:

Could you post the spectra from a couple of your LED light bulbs in your home, along with their ratings (power, "warmth", etc.)? That would be interesting to see...

From work, this type of light is used in some UK department stores. This is one is 4000K so slightly orange. All the LED SPDs I have seen have this typical blue spike and characteristic shape.

pinball1970 said:

This is one is 4000K so slightly orange.

Interesting. If you were to try to get good matches for two different 'strongly' coloured cyan fabrics (i.e. blues and greens reflected by the dye and a cyan dye for the other fabric) then you could risk seeing a mismatch when you get home or go outside where the illumination is uniform. Time was when people would take coloured clothes outside the shop to see what they would really look like. Those horrible flu tubes were very popular in department stored. Incredible spectra! I believe the M&S women staff used specially chosen makeup to make female customers feel inadequate and needing new clothes and makeup.

That spectrometer is a 'nice thing' to have at home but would cost a bit. I expect - I wonder; is the coloured bit of the display produced by a proper RGB screen or with printed mask with a monochrome display? Perhaps that's my age talking!

renault said:

yes indeed that would be interesting to see. I still don't have my question answered though - that is; the so called 'Edison' old style new LED bulbs clearly appear to show a yellowish light - and they claim a colour temperature of 1800°K. So how do you think they are doing this?

It is still not entirely clear what you are asking. Old style bulbs and LEDs have very different working principles.
That said, in both cases the light actually emitted will depend both on the "source of light" (the filament or the LEDs used) and the filters/coatings used.
Hence, there are many, many different ways to achieve a desired bulb temperature so I don't think there is one single answer to your question.
Note that there are also LED bulbs with tuneable temperature (from 1800-6500). Not sure how this is achieved, a best guess would be by using a combination of different LEDs

f95toli said:

Old style bulbs

We are not talking about hot tungsten filaments. We are talking of the (used to be) fancy filament-look-alikes which have a string of leds with extras- see the link in the post above.

f95toli said:

there are many, many different ways to achieve a desired bulb temperature

Not strictly true. The colour temperature achieved with a led device is artificial and a sort of match to a black body source. The low cost eastern manufacturers often think of a number ; a high number for blue-white and a low number for a warm white. In the old days, temperature meant temperature and was fairly easy to reproduce if you could rely on the voltage value.

f95toli said:

I don't think there is one single answer to your question.

you are absolutely right. All you can do is use (full price) led bulbs from just one manufacturer throughout a room or house if you want consistency. If you are into Feng shui then you have to use oil lamps.

renault said:

I am thinking about the health aspects of home led lighting and whether it is possible to filter out the shorter wavelength light. I am not sure what manufacturers do in order to produce their 'warm light' bulbs. They claim a lower colour temperature but how are they achieving that?

Worth reading this study.

https://www.ncbi.nlm.nih.gov/pmc/ar... exposure (especially,retinal cell damage [3].

In terms of kit some boxes come with Slightly opaque plastic covers to reduce intensity but the SPD I posted was with that filter.
I will post Lux info tomorrow.

Bottom line is that it is not a good idea to look directly into any bright light.

If you want to block UV, the usual material one uses is glass. So if you want your bulb to put out less UV - and this is mostly UV-A and the lowest energy half of that, so health impacts are at their smallest - the easiest and simplest thing to do is make the bulb thicker.

Theere's always the possibility of wearing tinted glasses in the evening. All levels of filtering are available and it could be cheaper than changing all bulbs in the home. It also can help with those loony blue headlamp bubs which everyone else seems to like.
I have some 'Cocoons' over-glasses for night driving and they are just the job.

Vanadium 50 said:

If you want to block UV, the usual material one uses is glass. So if you want your bulb to put out less UV - and this is mostly UV-A and the lowest energy half of that, so health impacts are at their smallest - the easiest and simplest thing to do is make the bulb thicker.

Virtually no UV in the commercial LEDs I've looked at. You can see that in the SPD I posted.

I am finding this thread very interesting. Thank you all for contributing. I think I would like to ask just two questions at this stage: 1) if it is possible to simply block out the lower wavelengths by 'just wear glasses at night' then why don't manufacturers envelop their bulbs with the same glass and stop the problem at source.?
2) If I understand correctly, manufacturers change the ratio of the elements which form the semiconductor in order that it produces light of a certain colour. So they can produce 'white' light. I guess then the question is how narrow a bandwidth of light can they achieve and can that be sufficient enough to eliminate all the lower wavelengths.

sophiecentaur said:

Interesting. If you were to try to get good matches for two different 'strongly' coloured cyan fabrics (i.e. blues and greens reflected by the dye and a cyan dye for the other fabric) then you could risk seeing a mismatch when you get home or go outside where the illumination is uniform. Time was when people would take coloured clothes outside the shop to see what they would really look like. Those horrible flu tubes were very popular in department stored. Incredible spectra! I believe the M&S women staff used specially chosen makeup to make female customers feel inadequate and needing new clothes and makeup.

That spectrometer is a 'nice thing' to have at home but would cost a bit. I expect - I wonder; is the coloured bit of the display produced by a proper RGB screen or with printed mask with a monochrome display? Perhaps that's my age talking!

yes I remember being told that the meat counter in supermarket was illuminated with lower colour temperature bulbs in order to make the meat look more red. It is true - look down the isle of the supermarket

sophiecentaur said:

Interesting. If you were to try to get good matches for two different 'strongly' coloured cyan fabrics (i.e. blues and greens reflected by the dye and a cyan dye for the other fabric) then you could risk seeing a mismatch when you get home or go outside where the illumination is uniform. Time was when people would take coloured clothes outside the shop to see what they would really look like. Those horrible flu tubes were very popular in department stored. Incredible spectra! I believe the M&S women staff used specially chosen makeup to make female customers feel inadequate and needing new clothes and makeup.

That spectrometer is a 'nice thing' to have at home but would cost a bit. I expect - I wonder; is the coloured bit of the display produced by a proper RGB screen or with printed mask with a monochrome display? Perhaps that's my age talking!

This switch to LEDs is a huge energy benefit as we have discussed before. M&S used TL84 not sure if they still do CCT of 4000K.
The thinking was walking around the store you would like to feel like a home atmosphere, so an orange light, rather than a bright blue light like D65 (6500)

Problem with that is the chromophores in the dyes and pigments behave differently in both lights, especially neutral colours where one colourant is not dominant.

Hence ladies taking products to the window to see what the “real” colour is.It is referred to as colour constancy, not to be confused with metamerism which is two samples matching in one light source but looking different in another.

LEDs will pose the same problem, one of which is a lack of UV content (worse than TL84)

That is an issue because anything where you are trying to showcase a bright white, will end up looking a little yellow.

This is because the optical brightening agent used in paper and textile substrates are activated by UV, absorbing that wavelength then throwing out visible light at the blue end.

Remove the UV you remove the effect.

I have never used a commercial UV LED but I think the CIE are adding two one or two to their “standard” illuminant list.

Also, sources of light that DO emit UV that could be dangerous are regulated (at least here in the UK), similar to the rules around lasers.
Not that this will stop people from buying UV diodes from China, but if you want to use a UV source in the workplace there are all sorts of H&S paperwork you need to do first (where I work we have dedicated courses for this).
Hence, as long as you buy from reputable suppliers there shouldn't be any health risks

renault said:

stop the problem at source.

What problem?

The fact that the UV-A emission is non-zero? That was true with incandescent bulbs. It's true even with no bulb.

IMaybe the amount of UV-A is unsafe. OK, but then you need to tell us what a safe level is and that the amount that is being produced. Without both numbers, it's hard to take a serious look at this statement.

You are always free to put bulbs in ghlass globes. Got a few in my house. Mostly for decorative reasons, but they are there. Why is this a problem for you?

pinball1970 said:

not to be confused with metamerism which is two samples matching in one light source but looking different in another.

Of course. You can make a 'white' illuminant that has massive holes in its spectrum which will match (metameric) a hot filament (black body). Narrow band pigments could appear almost black if they fall in the holes. It's important to distinguish between analysis of colours and synthesis.

pinball1970 said:

the optical brightening agent used in paper and textile substrates

Haha - to add further confusion to the discussion. In the context of regular colourimetry, the illuminant is assumed to be a black body. When it's not we have a different can of worms. I imagine the mqarketing of 'high quality' lighting (as used in surgery and printing) could risk being like the HiFi trade)

sophiecentaur said:

Of course. You can make a 'white' illuminant that has massive holes in its spectrum which will match (metameric) a hot filament (black body). Narrow band pigments could appear almost black if they fall in the holes. It's important to distinguish between analysis of colours and synthesis.

Haha - to add further confusion to the discussion. In the context of regular colourimetry, the illuminant is assumed to be a black body. When it's not we have a different can of worms. I imagine the mqarketing of 'high quality' lighting (as used in surgery and printing) could risk being like the HiFi trade)

Yes we have had a discussion on this too and that's when you demonstrated, reflectance data on a surface is not the same as discussing a black body.
I have learned a little since then but no where near your level.
I'm trying though!

renault said:

My understanding is that the shorter wavelengths (350-450nm.) destroy melatonin which leads to poor sleep and can result in macular degeneration.

As far as I am aware, there is no good evidence that blue and violet light has negative effects on people, including sleep disturbances and macular degeneration. There's an immense amount 'information' that gets tossed around when it comes to health, the vast majority of which is nonsense. Combine that with sensationalist news articles and people jumping on every new fad that pops up and you get an environment that highly susceptible to misinformation, fearmongering, and exaggeration.

renault said:

1) if it is possible to simply block out the lower wavelengths by 'just wear glasses at night' then why don't manufacturers envelop their bulbs with the same glass and stop the problem at source.?

Because then they wouldn't have white lights. They would have orangish lights.

renault said:

2) If I understand correctly, manufacturers change the ratio of the elements which form the semiconductor in order that it produces light of a certain colour. So they can produce 'white' light. I guess then the question is how narrow a bandwidth of light can they achieve and can that be sufficient enough to eliminate all the lower wavelengths.

You can't eliminate blue light if you want to achieve white light. You're going to need at least some.

thanks for your reply drakkith. The oft quoted sentiment that 'there is no good evidence' is not one I support. One can find and quote 'evidence' in every field which supports ones own preferred narrative. I prefer to take the cautious approach and for the moment will accept that low frequency light is able to cause harmful effects and if possible, without causing me too much inconvenience, I would like to take steps to avoid this in my home. It is this stance which prompted me to start this thread. Your second two points say that it is not possible to have white light without blue light. Just to be clear (I am learning a lot here !) would that be the same as saying that it is not possible to achieve white light from a source whose intensity is from around 500nm. to 700nm.?
Finally, is there a direct correlation between colour temperature and wavelength. I can't find a graph showing that there is.
cheers

renault said:

The oft quoted sentiment that 'there is no good evidence' is not one I support

That's your choice, but be aware that one can never prove the health effects are zero. If there were health effects for one person in a trillion, would we ever know? There aren't a trillion people. There have never been a trillion people.

And why near UV-A and/or blue light? Maybe its red light. Maybe it's green light. Maybe its light from incandescents. Maybe it only becomes safe if we chant the Ooga-Chocka introduction to Hooked on A Feeling. Once you detether your beliefs from the data, where does it stop?

renault said:

thanks for your reply drakkith. The oft quoted sentiment that 'there is no good evidence' is not one I support. One can find and quote 'evidence' in every field which supports ones own preferred narrative. I prefer to take the cautious approach and for the moment will accept that low frequency light is able to cause harmful effects and if possible, without causing me too much inconvenience, I would like to take steps to avoid this in my home.

You are free to believe whatever you'd like and to take whatever steps you'd like. But if you come here to PF and ask what steps you can take to protect yourself, you have to accept that we can only give you advice that is based on mainstream literature (or at least our understanding of it). Since that literature doesn't appear to clearly support the idea that blue light is harmful, nor that some protective measures like blue-light blocking glasses are even effective, you cannot fault us for saying so.

Note that you can find a correlation between almost any two things, so the fact that some studies find correlations between blue light and health issues isn't surprising. What is much, much more difficult is finding actual causation between the two.

renault said:

Your second two points say that it is not possible to have white light without blue light. Just to be clear (I am learning a lot here !) would that be the same as saying that it is not possible to achieve white light from a source whose intensity is from around 500nm. to 700nm.?

That's exactly right.

renault said:

Finally, is there a direct correlation between colour temperature and wavelength. I can't find a graph showing that there is.

No, but there is one between color temperature and color. The reason there isn't one between color temp and wavelength is that, on the whole, many different combinations of wavelengths can generate the same color. This is why the red, green, and blue pixels on your computer of phone screen can generate yellow, orange, and other non RGB colors.

Note that color temperature is based on a black body, which is a hypothetical perfectly emissive object that emits a predictable spectrum at specific temperature. For example, a 6,000 K black body looks white because its spectrum closely matches that of the Sun, which is the light source our eyes evolved to use. A 4,000 K black body looks slightly yellowish, and as we lower the temperature the color of the object becomes more orange, then red, then disappears completely once the temperature is too low to emit enough visible light to even see.

Also note that the descriptions used in describing lights, such as "natural white", "cool white", "daylight white", and the like are nothing more than marketing words and have little to do with the actual color of the light.

If you'd like to see the spectrum emitted by a black body at some temperature, you can use the following link to make your own graphs: https://www.spectralcalc.com/blackbody_calculator/blackbody.php

Drakkith said:

Also note that the descriptions used in describing lights, such as "natural white", "cool white", "daylight white", and the like are nothing more than marketing words and have little to do with the actual color of the light.

They are indeed marketing descriptions but they are not without validity. The CIE ( "Commission Internationale de l'éclairage" ) color response curves are attempts made in 1933 and 1964 to characterize a standard human's vision and color response. Within that context, the meaning of color is pretty well defined. All humans may not percieve colors this way but the majority will, and the system provides a useful, but not unique, characterization. So I quibble with disnissing it out of hand......

Thanks again for your reply - there is much to unpack there which I thank you for and will digest tomorrow. I must however take exception you your saying "You are free to believe whatever you'd like and to take whatever steps you'd like. But if you come here to PF and ask what steps you can take to protect yourself, you have to accept that we can only give you advice that is based on mainstream literature". This sounds like what the government has been telling us about the jab - 'that the science is settled and we must not be allowed to question it'. Perhaps there are people on this forum who might just not accept the consensus promulgated by 'mainstream literature'.

renault said:

This sounds like what the government has been telling us

Aye, the beauty of conspiracy theories is that the best conspiracies leave no evidence.

hutchphd said:

They are indeed marketing descriptions but they are not without validity. The CIE ( "Commission Internationale de l'éclairage" ) color response curves are attempts made in 1933 and 1964 to characterize a standard human's vision and color response. Within that context, the meaning of color is pretty well defined. All humans may not percieve colors this way but the majority will, and the system provides a useful, but not unique, characterization. So I quibble with disnissing it out of hand......

Perhaps I was overly harsh on the descriptions.

renault said:

This sounds like what the government has been telling us about the jab - 'that the science is settled and we must not be allowed to question it'. Perhaps there are people on this forum who might just not accept the consensus promulgated by 'mainstream literature'.

You have mistaken "there's no good evidence for this" with "everything is fine, stop questioning, you're being stupid". I don't care whether or not you continue to look into a topic. But I also know that health and safety information about any topic is almost always distorted and misunderstood by virtually everyone not professionally experienced with that topic. If I were to follow every piece of advice on every health topic I can find online I'd never eat red meat, never eat sugar, never eat fats, undercook my food, overcook my food, eat raw food, eat zero processed foods, not watch TV or use a computer, take only 'natural' medicine, get acupuncture or use healing crystals or something, not get vaccinated, not use sunscreen, not get an MRI or an X-Ray, drink non-fluorinated water, drink ionized water, not live near power lines, not use wi-fi or cell phones, use blue-light blockers... should I go on?

The only way people can function is:

1. Trust that mainstream science and medicine is generally correct.
or
2. Distrust mainstream science and medicine, but simply pick and choose what health issue they're going to get scared about and change in their lives since you literally cannot keep up with all the issues that crop up.

I choose to do number one and trust that modern medicine is generally correct and even when they are wrong they will tend to correct themselves over time.

We live the life of royalty, particularly in the USA and western Europe because our founding fathers embraced scientific method as fundamental truth and attempted to build a society on "self-evident truths" rather than superstitious fears. It has worked out pretty well for me so far, and I am apalled by any attempt to revert us to fearmongering. Those who would profit from it, in clerical robes or workaday garb, are far too sanguine in the attempt, and we need to be steadfast in understanding and demanding the level of proof dictated by scientific method. The balance is very tenuous and those who seed the wind will reap the whirlwind for us all.

Drakkith said:

Perhaps I was overly harsh on the descriptions.

You have mistaken "there's no good evidence for this" with "everything is fine, stop questioning, you're being stupid". I don't care whether or not you continue to look into a topic. But I also know that health and safety information about any topic is almost always distorted and misunderstood by virtually everyone not professionally experienced with that topic. If I were to follow every piece of advice on every health topic I can find online I'd never eat red meat, never eat sugar, never eat fats, undercook my food, overcook my food, eat raw food, eat zero processed foods, not watch TV or use a computer, take only 'natural' medicine, get acupuncture or use healing crystals or something, not get vaccinated, not use sunscreen, not get an MRI or an X-Ray, drink non-fluorinated water, drink ionized water, not live near power lines, not use wi-fi or cell phones, use blue-light blockers... should I go on?

The only way people can function is:

1. Trust that mainstream science and medicine is generally correct.
or
2. Distrust mainstream science and medicine, but simply pick and choose what health issue they're going to get scared about and change in their lives since you literally cannot keep up with all the issues that crop up.

I choose to do number one and trust that modern medicine is generally correct and even when they are wrong they will tend to correct themselves over time.

in the interest of not boring others, I will simply say that I wish you well with taking your 'number 1' choices, personally I tend to take your number 2 option, being guided by the 'follow the money' principal and generally what I see around me. I will be happy to continue this line of discussion if you want to create another topic if you wish, but now I am gong to bring this thread back on the topic I started.

renault said:

I tend to take your number 2 option, being guided by the 'follow the money' principal and generally what I see around me.

Fair enough but that is zero to do with how Science works and this is a scientific forum. I would read the Forum mission statement again.

"Our mission is to provide a place for people (whether students, professional scientists, or others interested in science) to learn and discuss science as it is currently generally understood and practiced by the professional scientific community."

Vanadium 50 said:

What problem?

The fact that the UV-A emission is non-zero? That was true with incandescent bulbs. It's true even with no bulb.

IMaybe the amount of UV-A is unsafe. OK, but then you need to tell us what a safe level is and that the amount that is being produced. Without both numbers, it's hard to take a serious look at this statement.

You are always free to put bulbs in ghlass globes. Got a few in my house. Mostly for decorative reasons, but they are there. Why is this a problem for you?

I am interested in your idea of putting bulbs in glass bowls. Just as an example, are you saying that if I put a jam jar over my bulb, the wavelengths of light emanating from the jam jar would change.?
I am also curious how the bulbs with variable colour temperature work. the claim variability between 2800 and 6000°K.