Led technology
- Thread starter Bronze
- Start date
Help Support Candle Power Flashlight Forum
Illum
Flashaholic
probably until quantum dot technology matures... not to mention the need to improve battery technology as well.
Drinking already this morning?
black kamagong
Newly Enlightened
hahaha c'mon bro be nice
id love to have a 3500 lumen pocket torch IF it can be done :devil:
jtr1962
Flashaholic
First off, this thread should probably be in fixed lighting if you're talking about screw-in light bulbs. Second, the honest answer is not until they get efficiency up to around 250 lm/W. You're limited to about 3-4 watts heat dissipation with the light bulb form factor. Active cooling really isn't feasible due to cost/reliability constraints. Right now it's pretty much impossible to get over ~600 lumens unless you go with an oversized heat sink. The only way to increase output will be when LEDs put out more light than heat. Given that Cree recently broke 200 lm/W we may not have as long to wait as I thought. Maybe in 5 or 6 years you'll see screw-base LED bulbs putting out the same as a 150-200 watt incandescent.
You might want to read this thread also.
Serial Chiller
Newly Enlightened
Wouldn't a peltier-element between the LED and the heat sink already do the job? No moving parts, no extra space and negligible extra cost. Sure, it would cause some more power consumption, but if the LED already takes like 30 watt or more, it wouldn't be that big a difference. Or maybe one day they come up with an LED that can sustain greater temperatures?!
etherealshade
Enlightened
This sort of thing has been discussed multiple times before. Regarding the use of a peltier heat pump in a portable lighting application, have a gander at this thread: Peltier cooled high power LED
They are horribly inefficient, and on top of that, they produce such an extraordinary amount of heat, you'd pretty much have to fan cool whatever heatsink you put on top of them.
Besides, it's totally overkill for pretty much any current LED application. There isn't any need to cool an LED below ambient room temperature, so you might as well save yourself the time and design a better heatsink in the first place.
They are horribly inefficient, and on top of that, they produce such an extraordinary amount of heat, you'd pretty much have to fan cool whatever heatsink you put on top of them.
Besides, it's totally overkill for pretty much any current LED application. There isn't any need to cool an LED below ambient room temperature, so you might as well save yourself the time and design a better heatsink in the first place.
jk037
Newly Enlightened
Suffice to say that a Peltier cooled LED would probably be capable of being driven to an impressive output, but thanks to the aforementioned inefficiency of the Peltier element this would not be practicable for any flashlight where size and battery life are significant factors.
DM51
Flashaholic
It is difficult to know whether this belongs in LED (emitters) or Fixed Lighting, but it certainly doesn't belong in LED Flashlights. I'm moving it to LED (emitters) for the time being; we'll see how it goes.
grazza
Newly Enlightened
I wonder if it will ever be possible for a LED, or other semiconductor, to only output radiation in the light spectrum and not heat.
Serial Chiller
Newly Enlightened
Of course, the LED doesn't have to get cooler than room temperature. But if you put a very strong LED into a normal flashlight and run it at maximum output for a longer time, the heatsink won't be able to eliminate enough of the heat. In this case, a Peltier element would make it easier to keep the LED at an acceptable temperature level. This would of course heat up the heatsink even more, but the hotter it gets, the more effective it gets. As long as it isn't actually melting, I don't see a problem with that.
It would probably be a good solution, if there was some kind of a thermostat that only activated the Peltier element when necessary. This would reduce the power consumption and avoid condensation.
TorchBoy
Flashlight Enthusiast
Yes, it can be done. You just need 8 XP-G running at 1.5 A each and if they're run in parallel, a battery that can give 12 A while maintaining a reasonable voltage. It would be easily pocketable. BUT you wouldn't need to worry about runtime of the battery because heat from all those LEDs would mean you wouldn't be able to maintain 3,500 lumens for more than a few seconds. :sigh: You didn't say you wanted lasting power.
blasterman
Flashlight Enthusiast
- Joined
- Jul 17, 2008
- Messages
- 1,802
You mean Cree actually putting 250lm/w emitters in production, or just issuing press releases saying they know how to do it? I'd say 80% of the people that frequent this forum don't know the rational difference (nor care)
And pray tell is there a problem with saying you know how to get a certain efficiency?You mean Cree actually putting 250lm/w emitters in production, or just issuing press releases saying they know how to do it? I'd say 80% of the people that frequent this forum don't know the rational difference (nor care)
LEDninja
Flashlight Enthusiast
luminus already has a 4*SST-90 module out.
http://www.luminus.com/stuff/conten..._cbm_360_w_product_datasheet_illumination.pdf
3000-4300 lm.
When people talk about fixed lighting bulbs they automatically limit themselves to the A-19 globe bulb which is 2-3/8" wide. A PAR-38 bulb is 4-3/4" wide giving 4X the cooling area. So if a A-19 can theoretically handle 600 lm, a PAR-38 can theoretically handle 2400 lm.
A 3000 lm incandescent bulb is 200 watt which is too bright for the typical table/floor lamp. It will probably go in a torchiere doing a ceiling bounce. A PAR 38 bulb will fit into a torchiere.
JTR1962 has commented in the past we should stop thinking bulbs but the entire fixture.
I look at the ceiling of my bedroom. The glass shade is 1 ft x 1 ft. How many LED lumens can it support heat-wise?
I look at the ceiling of an office. The linear fluorescent fixture is 2 ft x 4 ft. How many LED lumens can a 2 ft x 4 ft piece of aluminum heatsink?
A single light fixture that can produce 3000 lumens can be built now.
Getting the UL safety listing and the energy star efficiency certification for it is a whole different ball game.
http://www.luminus.com/stuff/conten..._cbm_360_w_product_datasheet_illumination.pdf
3000-4300 lm.
When people talk about fixed lighting bulbs they automatically limit themselves to the A-19 globe bulb which is 2-3/8" wide. A PAR-38 bulb is 4-3/4" wide giving 4X the cooling area. So if a A-19 can theoretically handle 600 lm, a PAR-38 can theoretically handle 2400 lm.
A 3000 lm incandescent bulb is 200 watt which is too bright for the typical table/floor lamp. It will probably go in a torchiere doing a ceiling bounce. A PAR 38 bulb will fit into a torchiere.
JTR1962 has commented in the past we should stop thinking bulbs but the entire fixture.
I look at the ceiling of my bedroom. The glass shade is 1 ft x 1 ft. How many LED lumens can it support heat-wise?
I look at the ceiling of an office. The linear fluorescent fixture is 2 ft x 4 ft. How many LED lumens can a 2 ft x 4 ft piece of aluminum heatsink?
A single light fixture that can produce 3000 lumens can be built now.
Getting the UL safety listing and the energy star efficiency certification for it is a whole different ball game.
jtr1962
Flashaholic
In theory yes, in practice no, and the only way to do it in theory is to color mix red, green, and blue LEDs to make white. Phosphor whites have inherent Stokes losses which would make them about 82% efficient in the best-case scenario ( that's with a 100% efficient blue emitter ). You can tweak the emitted spectrum to increase lumens per watt ( at the expense of color rendering ), but you still come up against the ~82% limit. All that being said, I think eventually we'll see LEDs which produce 75% light, 25% heat, perhaps even >80% light, <20% heat. Right now the best bin XP-Gs are up to 40% light, 60% heat. The best production blue emitters are about 50% light, 50% heat.
Maybe not Cree, but here's a 6000 lumen LED:
http://www.luminus.com/content1520
As far as the rest goes, well, the real question is, what are you using that much light from a single emitter module for?
http://www.luminus.com/content1520
As far as the rest goes, well, the real question is, what are you using that much light from a single emitter module for?
Basically, they already do. The energy that's not light is conducted and convected from the LED, not radiated. There's VERY little emitted radiation outside the visible spectrum.
