In 2024, why aren’t $3 60-year bulbs sold in every hardware store?
Did a modern type of Phoebus Cartel suppress that world-changing lighting innovation to protect their profits?
Next thing you know, a Big Pharma cartel is also suppressing cures for all sorts of diseases because marginally-effective medicines we need longer make better profits than cure-once medicines.
The article is speaking only about 5mm single diode LEDs, as far as I can tell. When you scale up LED packages, heat management becomes a huge issue for LED longevity, as do other elements like power distribution.
In a lot of LED systems now GaN LEDs are well used and not novel anymore. And we’re getting 100lm/W at 150,000 hour L70 values. But that’s just the LED. The driver electronics and such (which is small and cheaply made in budget consumer LED bulbs) are generally what fail first, rather than the LED itself.
Reeve said: @Vick
This, I have seen more Power Supply/Driver failures in integrated fixtures than failed LED chips.
You can take apart a dollar tree light bulb and look at the electronics. The AC to DC conversion is usually something super simple like half rectification, a capacitor, and enough LEDs in series to match the voltage from the capacitor. It’s basically a single white circuit board and it’s all enclosed in plastic. The heat dissipation is just some foil on the back side.
Luminaries made with LED in mind can easily last all those years and more, but this is where it gets tricky - the bulbs and the luminaries we tend to use them in have/had conflicting requirements - LEDs need to get rid of heat and conduct is away, luminaries that used conventional bulbs did everything to isolate the heat of the bulb.
A LED bulb now has to work in an environment designed for incandescent which limits how much heat it can generate and how much light it can produce, same goes for driver electronics.
Now the pricing - compared to incandescent, LED light is a rather involved product - LEDs, PCBs, PSU, heatsink, optics, and you can’t really separate those things apart - can’t have all of that for 3$, but still we got surprisingly close - now what was 60W bulb barely eats more than 7W and pays for the difference very fast.
When I’ve started designing LED lights right around 2009 - the LEDs back then could barely muster 50-70lm/W, quality of light was abysmal with CRI of 60-70, now you can buy bins with 230lm/w performance, CRI of 90 is normal, with 95 easily available, any color temperature you want is out there, 0.2W LED was something like 20ct a piece at 5k quantities, now a 0.5W LED is 2ct.
I had a client that needed as much light that can be served by 3 LEDs, but they went for 30pcs - why? because its that cheap.
@Niko
I wish everyone understood that and that the industry could more properly address the issue.
All that heat has to be removed right from the LED itself into some kind of heat sink and then ventilated… somewhere. i’m really struggling to find ceiling fixtures the provide the right LED bulb orientation and ventilation, and not just having a bare bulb sticking out.
It probably wasn’t as good as they made it seem in the article. Sometimes moving from a prototype to market is more difficult than people assume. They advertise a 100,000 hour run life for the bulb. Even the cheapest led you buy now has a 50,000 hour life. For context 50,000 run hours would be 8 hours a day, every day for 17.5 years. 80,000 hour bulbs and 100,000 hour bulbs are still available. Led technology has come a long way in the last 10 years. I suspect led’s will only get better with a longer run life as time goes on.
@Vince
not until the whole heat dispersion issue is resolved… Run an LED without overdriving it in open air and yeah, it will last on lifetime. Start over driving it and not dispersing the heat and it’ll last… months?
Most manufacturers of this sort of product don’t want to make products that last a really long time because it results in fewer sales. This has been a thing for a long time.
In theory, a LED will last a LONG time. If, proper power is given to it. However, that article does not talk about how the LED, that lasts 60 years, also has electronics that 1: don’t last that long, 2: will be skipped for cheaper manufacture, thereby killing the life of the entire UNIT for only 100hr. 3: yes, it’s called planned obsolescence. WHy would a company sell a “3$” bulb once? That is a bad biz model!!! You need people ot buy it again and again.
OH STFU, its not, its just bean counters squeezing every last cent out of it and forcing engineers to dance on the edge of whats possible.
Nobody sits at the R&D with the goal of ‘must last no more than 10k hours’, its more like ‘if we replace 125C rated capacitors for 11.5ct/pcs 120C rated ones for 10.75ct/pcs will it last about the same, eh probably, DO IT!’?
@Niko
You do realize there are more people involved with a product than R&D right? If you want to sell a 60 year light bulb that will never fail, go for it. You’ll make some good money this year. Maybe none next year, but…ya know, let’s leave it to the engineers to decide that.
@Jin
I do, and I’ve been in this industry for a long long time.
Nobody designs for 60 years, but for 50-100k hours, and we do test for it.
> You’ll make some good money this year.
Yeah, and then the next year, and then this office block does a redesign, and then we sell to the manufacturing plant that replaces its retrofit lights with better ones, oh then to this chain of supermarkets builds a logistics center, and then our sales establishes distribution to Arstotzka, and then and then and then and then.
And somehow not a single time I designed for a fixed amount of hours, funny that.