The biggest problem of LED bulbs is usually heat. The design is just not suitable for LED, it would be great if we get a new standard.
First thing is voltage, 120/230V is not suitable for LEDs, if lamps would have an external 12V or 3V power supply a few centimetres away from the LED, this would be great.
The second thing is, that lamps need some kind of (passive!) cooling solution, to remove the heat from the chip. A lot of lamps are more or less closed and heat up as a whole. This is really bad for all the electronics, and probably the main reason why LEDs fail early.
Aren't filament bulbs the next evolution for both heat and power supply? The filament design puts the heat generation away from the electronics, and the electronics themselves are simpler because each filament can run off line voltage (or something along those lines). But they have more of a flicker problem IIUC.
The nice thing about newer filament bulbs is you can get 1600 lumen bulbs that are rated for enclosed fixtures now.
Yah, filament bulbs are the future of LEDs they solve basically all the existing problems of LEDs - heat, point-source lights, directionality, driver-size. And they are cheaper.
I believe they are referring to LED filament bulbs, which are designed to look like traditional filaments but have the benefit of separating the LEDs from the electronics.
It does seem that the filament bulbs check a lot of boxes. The driver is so small it fits in the socket itself. If they can put 1600 lumens in an enclosed fixture, it does suggest they do better with heat. My impression from buying a box is that these are a next generation bulb design, not quite holy grail but a significant step. No more buying bulbs where the bottom half the globe is plastic housing a bunch of driver electronics.
Without some kind of driver, the light flickers strongly at 120Hz. For sensitive people, the effect is extremely unpleasant.
One can get perfectly nice drivers and perfectly nice LEDs for less than $25 all in. Sure, many people prefer paying less for a screw-base lamp, but the result is not as good.
LED filament bulbs do indeed use a driver circuit. Some use a voltage above mains [1] while others drive the LED chain around mains voltage but in a smoother regulated way [2].
Sometimes there's just a basic dropper cap circuit, though. Those flicker but not too badly. The LED design in these is different than in a discrete LED, with the phosphor being quite thick and having a distinct slow fade/afterglow when switched off. I think in this case they basically use the phosphor to smooth out flickering instead of electronics.
The efficiency difference in a power supply running off 120V or 12V is not significant. The power supply is a negligible portion of heat output. They would last longer if separate though. Slim recessed LED fixtures have their power supply in a separate box.
I thought most household bulbs chopped it down to something like 30-40V.
Some lower spec lighting devices might take 120V and run LEDs in both series and parallel, but as LEDs fail, the rest get more current, leading to more failure… positive feedback loops for the win!
the parallel chains of serial LEDs are balanced during production so they all consume the same current and if any LED in a serial chain fails, they all stop working, simply eliminating that chain from the circuit. Of course in reality it's not that simple.
It’s not really very good to use them like that though - without a huge cap you’ll get pretty prominent flicker - I expect it’s probably just cheaper to stick in a cheap SMPS at that point if that’s something you care about.
The LED strip behind my TV is powered by the USB port on my TV. Most longer strips (5m-20m) use one of a small number of form factor strips and 12V or 24V supply. There are also quite a few USB powered lamps, but a bulb form factor would be interesting.
Whole-house 12/24VDC power wiring has been a thing for at least 50 years, specifically intended for lighting. It's never proven widely popular, but it's certainly a realistic option. Probably more so in new construction than retrofitting into existing buildings, though.
Whole-home PoE for lighting might be another interesting approach!
In either case, you'd need a very beefy DC power supply somewhere to get enough current, or individual power supplies scattered inside walls, etc. which sounds like a maintenance nightmare.
Or keep the voltage high and step down with a buck converter in each socket as needed and negotiated with the device: there's a good argument to use ~300VDC as a future home electrics design. High-power devices (water heaters etc) will work better and more efficiently with DC, and low-power devices can deal with 48VDC with a lightweight step-down converter (5A buck converters are tiny)
From a purely technical standpoint, I agree completely. However, given the debacle around USB-C cables & PD, I'd be very worried about a world where I have to trust devices to properly handle such high voltages.
I've been derating leds for a while. no experimental log, but at 90% of input power they really don't seem to degrade at all over periods of many years, and they are a lot easier to thermally manage.
What's the context here? Are you somehow reducing power to standard 120/240v LED lightbulbs (how)? LED light strips? Custom designs using raw LED modules?
iirc (and I may not be remembering correctly), heat is what causes the degradation. Cutting power makes the LED run cooler.
For personal use at home it's important to buy high quality lightbulbs for this reason. Cheap ones will cut costs on cooling and fail prematurely or experience degradation if they don't fail. Better quality bulbs will last longer with less degradation. It can be hard to evaluate what's good; lots of companies will sell you cheap crap at inflated prices. Ikea uses solid metal bases on many of their bulbs which helps with heat (they use good CRI emitters too). I suspect weight might be a good heuristic to evaluate quality. Also avoid the high power ~1600 lumen ones, it's common for them to fail early even for the higher end ones, they just make too much heat to dissipate in the standard bulb form factor. Afaik Ikea eventually discontinued their 1600lm product, probably for this reason.
Cool! Is this a commercial product that we can buy?
I worked on a similar system a while ago for personal use. 300 watts of BLX COBs. Prototypes were good but they ran very hot (105 degrees) with the cooling system I had on them. Improving cooling and refining the design past the crude prototype would have made it too expensive to be practical for me, or for the market if I chose to sell it.
I'm probably going to be building a new house in a few years, it'd be nice if I could run 48v wiring throughout the house with standardized connectors and control modules. The age of 120/240v with screw socket bases needs to end. But I doubt the industry will have standardized on a DC system by then.
I wonder if there are any actively cooled fixtures on the market for standard lightbulbs. The common recessed in-ceiling fixtures are not great for thermals, the closed top blocks natural convection and traps hot air. A slow (silent) fan should do a lot to help with bulb longevity. I might end up doing a custom design for my house if there's nothing on the market.
not really. we used to sell them, but its a bit fussy. people had quite legitimate concerns about what to do when the modules went. margins are good, but not as high as fine art, and the labor is about the same for us.
since our primary activity is metal sculpture....we really don't have to deal with cooling issues. 10-20 pounds of metal, even steel, makes a _really good_ heat sink, even if it is isn't optimized for maximum convection. and now your light elements are little 3cm circles a couple mm high, and you can tuck them wherever you want.
it wouldn't be too hard to make you own standard. the constant current drivers often have pretty wide input ranges, especially if you can stick with simple bucks. the part where I really kinda fell down was connectors oddly enough. 3 pins, moderate current, few cycles, low cost, high contact reliabilty. anyone with better connector-fu than me have a suggestion?
USB-C might work. USB PD supports 5a at up to 20v, looks like there's also a higher power 50v 5a spec coming up. But if you're using custom PCBs on both ends you don't really care about the PD specification, as long as you're using good quality cables you can treat the connector/cable as a dumb wire that supports sending 5 amps down a few pins. The data lines may also be useful for sending back data from the emitter module, if you had any need for that.
Desktop computers use the Molex Mini-Fit Jr connector for ATX and PCIe power connectors. 6-pin PCIe power cables should be easy to source on the market, you wouldn't need to get custom cables manufactured. Those can carry ~150w. Might also be possible to cheaply source cables with just a 2 or 4 pin connector?
3 single Anderson Powerpole connectors (they are designed to slide together, and a drop of CA glue can make the connection permanent). Available up to 45A
USB? -48v? And there's a whole ecosystem of 12v stuff under the RV umbrella.
Seems like what's lacking isn't a standard (or rather the basis for a standard), but a market leader to say "we're gonna back this". But you've got way more insight than I do and I'm undoubtedly underestimating the effort.
From what I've observed, I mostly attribute bulbs dying to the power electronics rather than the LEDs themselves. So unless they've beefed up the driver circuit on these bulbs, I would think they'd fail just as often?
Yeah that's a very common failure mode. Like the LED itself, the power supply electronics are also intolerant to heat (see my comment on parent). So all the same principles apply. Companies which cheap out will use barely-adequate components in the power supply circuitry and skimp on cooling. The power supply will run hot and fail after a few months leading to total failure of the bulb, or if it manages to keep limping on, the LED will have noticeably degraded within a few years due to the heat.
Reducing power will help a lot with this, less current flowing through the power supply and less heat generated.
I've primarily seen driver failures. from some vendors, really quite smelly and spectacular ones. but it does seem to correlate pretty strongly with particular vendors or designs.
In related news the Bay Bridge in San Francisco will have lights out permanently from this weekend due to high maintenance costs of the LED bulbs. I wonder if they could just lower the brightness of the bulbs.
> “The hardware is failing at a rate faster than we can cost-effectively keep up,” Ben Davis, the founder of Illuminate, told KRON4 in January.
> “It’s a challenge of the current lights. They’re good but they have to exist in an area that has rain, fog, high winds, grime 24/7, vibration, expansion and contraction,” Davis said.
This has nothing to do with brightness or with the bulbs being LEDs. Their hardware is not properly suited for the environment it's being used in. If you put an Arduino in your car, don't be surprised when the non-automotive-grade electronics fail after a few years. The EEs must have not done a great job when initially designing this light project. From to the project's website (linked by your article):
> They must be replaced with a new system that is custom engineered to perform in the harsh environmental conditions of San Francisco Bay.
In their defense, the original system was only supposed to be up for one maintenance cycle, but it was extended due to the massive popularity of the installation.
> must be replaced with a new system that is custom engineered [for] San Francisco Bay.
Other major port-cities around the world, with comparable climate conditions, have presumably transitioned to LED lighting over the past decade - why can't they save taxpayer's money by simply buying-up whatever Valencia or Singapore is using instead of having their own NIH Syndrome and insisting an entirely bespoke system just-for-them be used?
This is a decorative/art installation, not the bridge's engineered lighting. So whatever Singapore is doing is not relevant (except if it's also for similar purposes).
Did you even read the article? This is a public art installation funded by a non-profit. Not taxpayers. The installation was intended to be temporary and has already gone way past it's design lifetime.
This is so weird though, because the lights that line the catenary cable are older and not proposed to be turned off (and anecdotally are much less likely to be broken on a given night).
Maybe it's the scale, relative inaccessibility, or remote control that's the problem.
Honestly, I never loved the 'light show' effect anyway - maybe they should just light the suspender cables with normal lights and call it a day.
These lights were originally a public art exhibit that was supposed to be for two years, and then was extended. [1] They aren't used for anything essential.
There's absolutely no reason they won't last if properly engineered. We have a very good understanding of how and why LEDs fail over time.
If the lights need to last 10 years, then the mechanical & electrical engineers can make a design that will last 10 years. That's not hard to do. The reason many LEDs, especially consumer LEDs, fail early is due to excessive cost cutting.
Man, these guys are always asking for more money. They want to take the cash and expand the installation. I really like the lights, but come on, you can't keep adding maintenance. It's a bad idea.
Other reviews of these things say they aren't especially bright and have a very mediocre CRI. It's not clear that anybody outside of Dubai would want these anyway.
Reminds me of Homer Simpson with a monocle sitting in a car twisting the steering wheel..."How does this compare with a train, which I can also afford?"
Think about it from the point of view of a manufacturer/distributor. Speculating a little:
Suppose you decided to carry them and they cost more, both from a material perspective and a reduced repeat customer perspective. Cheap competitors in stores won't just go away and they'll be on shelves next to them. Brand recognition and/or loyalty will only go so far and way too many people will prefer the cheaper cost. So the manufacturer/distributor have to provide both or risk losing sales. Now there's an extra sku in the mix and I'd wager that even with brand recognition the cheaper ones will outsell the Dubai ones, which makes the distribution overheads even more unpalatable. And so on.
I'm sure they've done their research and it doesn't add up from a business sense. It works in Dubai because the regulations eliminate part of the race to zero.
Even selling them directly online is problematic because of distributor territory, exclusivity etc. If the manufacturer competes with its distributor network (that they rely on) then that doesn't end well either.
I doubt it's any sort of conspiracy by 'big led' to keep you from buying them, but is more to do with mundane and unfortunate things like above.
They're under contract, but just as the "make them and we'll keep others from competing with you here", not "you can't sell these elsewhere".
The only reason you "can't buy them" is because you wouldn't (or at least people wouldn't). You can't easily tell how good the LED setup is in the bulbs, and price isn't necessarily a good indicator, so you're unlikely to pick the more expensive item.
No that's a different lamp, but Clive did a teardown of that one too.[0]
Compared to the Dubai lamp it still drives the LED filaments at 0.25 W each, but the driver circuitry is traditional and doesn't have the unusual features from the Dubai lamp. These include component-level redundancy and automatic brightness compensation (this breaks dimmers, so architects hate the bulb mandate).
Nothing ruins a nation's (carefully cultivated) high-tech luxury image like all the bulbs dimming simultaneously with the grid voltage!
The power incoming to a home is 220/240 in the USA. You just split the phase at the mains box. I have several 220V-native light fixtures in my kitchen.
bigclivedotcom: The lamps you're not allowed to have. Exploring the Dubai lamps: https://www.youtube.com/watch?v=klaJqofCsu4