The big problem for me is, as a consumer, how do I know that brand X is producing quality bulbs that'll last a long time?
The issue is that LED bulbs aren’t simple devices like incandescent bulbs. LED bulbs have an electronic power supply inside which drives the LEDs at a constant current.
Power supply design is a major subfield of electronics engineering and there are all kinds of tradeoffs you can make to optimize for different goals. Consumer electronics almost always optimizes for cost, to the detriment of all else.
It is possible (and not very difficult) to design LED bulbs that will practically outlive their owners [1]. The problem is that it requires putting more LEDs in the bulb and driving them at lower current. This makes the bulb cost more and the only benefit is longer life. For a manufacturer, there are nothing but downsides to this approach.
>It is possible (and not very difficult) to design LED bulbs that will practically outlive their owners
It is also possible (and not very difficult) to design incandescent bulbs that will outlive their owners. In fact, the first mass produced light bulbs generally lasted 2,500+ hours. In the 1920s, the major bulb manufacturers formed the 'Pheobus Cartel' in Geneva and secretly colluded to limit the lifespan of bulbs to 1,000 hours to boost sales [1]. Another example of planned obsolescence harming consumers and the environment.
> On 23 December 1924, a group of leading international businessmen gathered in Geneva for a meeting that would alter the world for decades to come. Present were top representatives from all the major lightbulb manufacturers, including Germany’s Osram, the Netherlands’ Philips, France’s Compagnie des Lampes, and the United States’ General Electric. As revelers hung Christmas lights elsewhere in the city, the group founded the Phoebus cartel, a supervisory body that would carve up the worldwide incandescent lightbulb market, with each national and regional zone assigned its own manufacturers and production quotas. It was the first cartel in history to enjoy a truly global reach.
An also should add that color temperature on incandescent lamps play a role on its lifespan, want long lasting lamps? Lower the current (or increase the resistance).
incandescent light bulbs efficiency increases with their temperature/current. At low enough current they will last long enough but waste a lot energy as well.
You can dim them, and provide a slow start to prevent the inrush current (which is like 10times more than nominal with tungsten resistance increasing due so high 2500K temps).
Thankfully, LED lighting will probably be gone within 20 years, while incandescent will be coming back more efficient than LED could dream. Already bumping up against theoretical maximum efficiency, LED lighting can't get any more efficient, and the better LED is at color rendition, the less efficient it is. But there are vast amounts of improvement available for incandescent lighting, and a group at MIT has already created incandescent light that is twice as efficient as LED.[1]
> Exciting, but that article is from 2016 and here we are in 2023 with no commercial availability.
That is a fantastic point. If they can't revolutionizing lighting in 7 years, it will never happen. Oh, btw, LED was invented in 1962, but it only took about half a century for the commercial viability of high powered LED lighting to appear and begin to take over the lighting market in 2011.
The concept is interesting, and reading the article reversed my initial response to treat this as a crank concept.
That said ...
... LEDs involve finding materials which exhibit specific quantum behaviours which correspond to human visual acuity.
Photonic bulbs involve the much simpler blackbody radiation concepts of not only incandescent light bulbs but hundreds to thousands of millennia of previous experience with combustion-based lighting ... and, yes, the added twist of finding viable IR-reflective / visual-spectrum emissive materials.
The second problem seems more reasonably simpler. One would hope that progress might be occuring at a more rapid rate.
(There's a similar argument I've used to contrast nuclear fission, which was commercially exploited within two decades of first demonstration, and nuclear fusion, which coming on a century from its theoretical understanding remains not even experimentally demonstrable on a continuous, energy-positive basis, let alone in commercial application. Some problems are just hard.
Yes, there are thresholds and breakthroughs, and they do occur. But given a few decades of lived experience matching advertisement to delivery, as well as a stronger awareness of historical examples and trends, patterns do become evident.
And that said: I will keep an eye on this. It does have the advantages of being simple, based on very well-proved technology, and a reasonable extension of same.
Most of the capital invested in incandescent technology was related to the Phoebus cartel. The pressure to make incandescent efficient was just never there until very recently. LED has severe problems, some solvable, some not. The article mentions flicker, but really should have specified. Nearly all LED drivers employ PWM. There are constant current LED drivers, and they are more efficient than PWM drivers, but PWM drivers are cheaper to design and manufacture. While many will claim PWM doesn't bother them because they can't detect it, they're not only exhibiting callousness for those that are bothered and harmed by PWM, they're falling into a fallacious trap, i.e. what they don't know and can't detect can't harm them, which is patently false, and one counter example is carbon monoxide. PWM LED drivers are now ubiquitous, and it's effects range from annoying to painful, as anyone that has experienced migraine can attest. Regarding the actual light LED produces, nearly all LED available are weighted towards the blue spectrum, and this light has been shown to massively mess with wildlife and shorten human lifespans by years by messing with circadian rhythms, which strangely can lead to diabetes and heart disease. LED proponents are obsessed with brightness, but this is also a trap, because brightness is not as important as what can be seen. Consensus among lighting and eye experts is that more can be seen with a dimmer light that reproduces color perfectly than with a much brighter light that does not. As we mitigate these issues with LED, the drivers become more expensive and the LEDs become less efficient. And as the article mentions, the phosphors of better color-producing LEDs will fade rather quickly and over time no longer reproduce colors accurately. And best LED can ever achieve is to perfectly match what incan does with little to no development, What it looks like, if the trend can be detected, is that as we fully mitigate the problems with LED, the more expensive it becomes until its cost and durability nearly reaches parity with what we already had with incandescent. It's becoming a wash.
Most of the capital invested in incandescent technology was related to the Phoebus cartel. The pressure to make incandescent efficient was just never there until very recently.
LED has severe problems, some solvable, some not. The article mentions flicker, but really should have specified. Nearly all LED drivers employ PWM. There are constant current LED drivers, and they are more efficient than PWM drivers, but PWM drivers are cheaper to design and manufacture. While many will claim PWM doesn't bother them because they can't detect it, they're not only exhibiting callousness for those that are bothered and harmed by PWM, they're falling into a fallacious trap, i.e. what they don't know and can't detect can't harm them, which is patently false, and one counter example is carbon monoxide. PWM LED drivers are now ubiquitous, and its effects range from annoying to painful, as anyone that has experienced migraine can attest.
Regarding the actual light LED produces, nearly all LED available are weighted towards the blue spectrum, and this light has been shown to massively mess with wildlife and shorten human lifespans by years by disrupting circadian rhythm, which strangely can lead to diabetes and heart disease.
LED proponents are obsessed with brightness, but this is also a trap, because brightness is not as important as what can be seen. Consensus among lighting and eye experts is that more can be seen with a dimmer light that reproduces color perfectly than with a much brighter light that does not.
As these issues with LED are mitigated, the drivers become more expensive and the LEDs become less efficient. And as the article mentions, the phosphors of better color-producing LEDs will fade rather quickly and over time no longer reproduce colors accurately. And let's realize that the best LED can ever achieve, what the technology has always been striving for, is to perfectly match what incan does with little to no development. Maybe someday LED light will perfectly match incan light, but it is not today and it isn't next year.
What it starts to look like, if the trend can be detected, is that as we fully mitigate the problems of LED, the more expensive it becomes until its cost and durability nearly reaches parity with what we already have with incandescent. It's starting to become a wash.
So unlike LED, if incan can be made more efficient, and there is a massive amount of room for improvement in efficiency there, then incandescent undoubtedly will return and dominate the lighting market. I liberally estimated 20 years, but it may take longer, but it really doesn't matter to my major point, which is that incandescent is coming back, and this is a very very good thing, because LED light, as efficient as it is, still absolutely sucks.
It seems like this could be done differently, and perhaps more cost-effectively. Can't give cites right now, but here's a path I'd explore if I were in the field.
I'd pattern the inner surface of the glass envelope with a cube texture - think of taking a cube and pressing a corner normally into a clay surface, then removing the cube. This pattern is a so-called corner reflector, and returns incident light to its source. Figure the cube indentations at about 0.5mm deep, close packed. I'd deposit a dielectric film reflector stack tuned to reflect most infrared radiation onto this surface.
This combination would transmit visible light, but would reflect IR directly back to the filament, reducing the amount of electrical power needed to maintain filament temperature. Glass textural molding and dielectric film deposition are mature technologies. I think this could readily triple incandescent lamp power efficiency, maybe even better.
Perhaps the planned obsolescence helped the consumer, because perhaps there would have been no willing producers if producing the lightbulbs at a price the users were willing to pay for wasn't going to turn out to be profitable, with respect to setting up production in the first place and then producing until the investment was paid back.
When we're talking market price, we have to acknowledge that it is a meeting of the price needed to bring a product to market and the price the consumer is willing to pay. We can't assume that the price of the longer lasting bulb would have been attractive to consumers, when compared to the price of the shorter-lived bulb, even if they had all the information available.
It's perfectly valid for a person to decide they'll spend more over the long run, rather than ponying up a larger sum now. And it's perfectly valid for producers to take the chance of deciding this for the consumer. As Henry Ford noted, "If I had asked people what they wanted, they would have said faster horses."
Was anybody stopping anyone from offering the consumer a higher-priced and longer-lasting bulb?
Yes, but this is exactly what I'm asking. How do I find the ones designed in a buy it for life way? Or at least ones designed to last longer than the crap on the Home Depot shelves.
You can't buy a bulb that is rated to last a lifetime, but you can buy ones rated to last for 10+ years if used 8 hours a day (i.e. 35,000 life hours).
I retrofitted entire house - 200+ bulbs and fixture retrofits more that 5 years ago. I had one or failures since. I bought highest CRI bulbs, i.e. most expensive, and they work well. (Also, do not use bulbs for downlights - get entire "led can light fixture retrofit")
I used 1000bulbs.com because I can filter/read specs there, but you can get specs for any high-tier vendor and buy elsewhere.
Just another thumbs up for 1000bulbs. I've made a small order through there, but I like my bulbs. It seems to get very expensive quickly going beyond 93 CRI.
If you don’t trust the manufacturers, you’ll have to find an expert to trust, or tear down the bulbs and examine them yourself. The best way to know for sure is to measure the current being driven through the LEDs and compare it to their maximum rated current. The ones that don’t last long are usually being run at their maximum current, producing a lot of excess heat which shortens their lives dramatically.
I don't think it's untrustworthy in a sense you are deliberately lied to. I think more likely problem is that there is quality control issues and entire batch of particular bulb is compromised, like can happen with anything else - hard drives, RAM, pencils.
In general it's quite reasonable. Cheaper bulbs failed on me, more expensive ones work just fine for many years.
One thing is I wonder about if phosphorus (or whatever chemical they use) is burning out over the years. I.d. do I get worse light quality as these bulbs age?
I imagine light consumption could be regulated like food could be. As a smoker paying about 20 times more than 40 years ago I feel entitled to say that approach was a good idea.
I've noticed that LED bulb packages say they'll last several years, but so many of them die after just a year or so. I should start tracking this precisely, but I've just gotten the sense that they often die prematurely.
Assuming that the retailer can actually get a refund or replacement from the manufacturer (for some products, only the consumer can actually do that), usually they just replace the item or the wholesale price. They don't replace the retailer's gross margin. Thus the retailer ends up eating the cost of things like stocking and actually processing the warranty. For e.g. a one-person operation, that means extra work for zero profit.
It's usually not that much, but I can see why they might eventually be upset about it.
> The issue is that LED bulbs aren’t simple devices like incandescent bulbs. LED bulbs have an electronic power supply inside which drives the LEDs at a constant current.
Not all of them! I was very surprised to open up my generic outdoor patio LED bulbs and find two strips of LED filament wired directly to power.
AFAICT it’s just enough LEDs in serial for 120VAC at 60 Hz to be “good enough” that they survive for “long enough”.
The issue is that LED bulbs aren’t simple devices like incandescent bulbs. LED bulbs have an electronic power supply inside which drives the LEDs at a constant current.
Power supply design is a major subfield of electronics engineering and there are all kinds of tradeoffs you can make to optimize for different goals. Consumer electronics almost always optimizes for cost, to the detriment of all else.
It is possible (and not very difficult) to design LED bulbs that will practically outlive their owners [1]. The problem is that it requires putting more LEDs in the bulb and driving them at lower current. This makes the bulb cost more and the only benefit is longer life. For a manufacturer, there are nothing but downsides to this approach.
[1] https://youtu.be/klaJqofCsu4