Theoretical brightness limit of a LED?

[Argon]

First off hi all found this site about a month ago and have only just now decided to sign up. Have also received a lodc-ce (soon to be mated with a 10440:naughty and a l2d-ce. So far I'm extremely happy with these lights.



Anyway with all these new brighter led coming out every few months I was wondering if there was a theoretical max brightness that a single led will reach? Or will a led one day out perform a HID. This is in regard to flashlights. I have searched plenty of times but can't seem to find the answer I'm looking for. Can anybody help with this one?

 

[carpdiem]

Re: Thererotical brightness limit of led's?Theoretical brightness limit of a LED?

Well, yes and no.

There's certainly a theoretical limit on efficiency. That's simply saying that you can never generate more energy of light than you put in. I.E.- the maximum possible efficiency is 683 lm / watt (see the excellent article at http://en.wikipedia.org/wiki/Luminous_efficacy).

However, you can always put more watts of power through something to raise the net power output. This becomes a question of energy storage (i.e.- how awesome can we make our batteries?). In this sense, the only theoretical limits are probably on our ability to store antimatter, which really is so far into the future that it's not really worth discussing as a limit.

 

[TMorita]

This is like asking, "What is the maximum theoretical water capacity of a single water tank?"

Toshi

 

[Derek Dean]

Welcome to CPF Argon. Lot's of nice folks here, and cool torches too!

 

[2xTrinity]

There's certainly a theoretical limit on efficiency. That's simply saying that you can never generate more energy of light than you put in. I.E.- the maximum possible efficiency is 683 lm / watt (see the excellent article at http://en.wikipedia.org/wiki/Luminous_efficacy).

However, you can always put more watts of power through something to raise the net power output. This becomes a question of energy storage (i.e.- how awesome can we make our batteries?). In this sense, the only theoretical limits are probably on our ability to store antimatter, which really is so far into the future that it's not really worth discussing as a limit.

Right now the fundamental limitation on LED power (not efficiency) is heat buildup -- LEDs are a very sensitive to heatbuildup, and anytime you have high amount of power dissipated in a very small area (a couple square millimeters) it's very easy to reach extremely high temperatures. This is why heatsinking is so important for good performance now.

One option is just to use a wider die, or multiple parallel dice, but the problem there is that with a larger surface, your ability to project the light into a beam for use as a flashlight goes down -- since your die is less and less like a point source. Diffused light at this point is easy -- fluorescent tubes put out a good 100 lumens per watt of diffused light -- it's the ability to focus LEDs that really sets them apart.

However, what I expect to see in the short term is continued improvements in efficiency. However, since overall radiant efficiency (power in vs power out) has double from say around 10 percent to 20 percent -- so the amount of light you get for a certain power is a lot more, but you still get close to the same amount of waste heat to deal with. Eventually as we approach maximum efficiency, incremental improvements in lumens per watt will be small, but improvements in heat buildup will be big

-- say we improve from 80% radiant efficiency to 90%. The increase in apparent brightness won't be very much, but heat buildup will all of a sudden be cut in half. That means you could run double the power and still be running at the same temperature, or concentrate the power into a smaller die and be able to achieve a tighter focus.

 

[luminari]

I've seen this question before on CPF, but still, it's always interesting. Seems like there's still a long way to go up.

It's interesting that compact fluorescent bulbs have much lower efficiency than the big shop tubes. Still, I dislike them all I can't wait for the day that those flickering, mercury-filled monsters get killed by LEDs.

 

[Calina]

Actually the quantity of mercury in there is minimal. Nevertheless, I do not like fluorescent lights that much either.

 

[2xTrinity]

I've seen this question before on CPF, but still, it's always interesting. Seems like there's still a long way to go up.

It's interesting that compact fluorescent bulbs have much lower efficiency than the big shop tubes. Still, I dislike them all I can't wait for the day that those flickering, mercury-filled monsters get killed by LEDs.

The more efficient tubes and CFLs for that matter don't flicker, since are run by electronic ballast that switches the current at 20kHz (absolutely undetectable), rather than inefficient magnetic ballasts. The difference between the two is huge -- we just replaced the shoplight in our garage from an old 2x40W T12 system, altogether putting out something like 60 lumens per watt, with an inefficient reflector, CRI of around 60, and noticeable flicker. Now we have an electronic T8 fixture with a much more efficient reflector, 2x32W power consumption, and 86 CRI. Probably less mercury in the modern tubes by a factor of more than ten. It made a night and day difference (about double the light, much nicer light quality, no more irritating flicker), and since then we've been putting in those fixtures in other rooms as well. In one room, we replaced a 4x50W track light system with the T8s in an open fixture -- about double the light from a third the power, and no more irritating shadows cast all over the place.

The CFLs give up efficiency because the self-contained ballasts are cheap/inefficient compared to good external ballasts. Also, the coiled shape is inefficient for distributing the light as a lot tends to bounce back into the tubes, or into the base. Finally, the ones that are desigend to be incandescent "look alikes" at 2700k are all terrible in my opinion, they sacrifice a lot of color rendering to get that style -- going up to 3500K (similar look to filtered incandescent) is much better.

Also, in a technical sense, white LEDs actually are fluorescent lights -- they use a phosphor to convert blue light into other colors. Since they are point-sources though, they can work in a lot of places that fluorescent can't -- such as if you want spotlighting, or low-lighting (fluorescent works best for bright area lighting). They can also be dimmed and cycled on and off without affecting lifespan -- in fact efficiency will generally improve when the lamp is dimmed.