Brightest red LED @350mA?

[Cemoi]

The red Luxeon Star has a claimed flux = 44 lm @350 mA.
The Cree XR-C is 39.8 lm @ 350 mA.

Any brighter red LED @ 350 mA?

 

[knabsol]

You also have the Seoul P4 with 48lm @ 350mA (625nm).

 

[Lightingguy321]

Keep in mind those numbers are the average outputs. Each red LED from a specific reel will have a different lumen output. For instance with Luxeon LEDs they can be any where between Q bin (~30-45 lm), R bin (~46-50lm) and S bin (max around 70lm). Typically you are probably going to be looking at a Q or R bin Red LED, S bins are rare across the board for Luxeon I and III. As for Cree and Seoul I am not as knowledgable but the idea of different flux bins applies to those 2 companies as well.

 

[Oznog]

Luxeon Rebel Red is 40 lumens, Red-Orange is 50 lumens.

The red types don't perform as well when you try to up the current, BTW. The AlInGaP technology they use has a high efficiency loss with elevated die temp than the white/blue/green InGaN tech AND it a higher thermal resistance which raises the die heat.

So if you wanna push for power that's something to keep in mind. I liked the Rebels because you can stack a lot of them in a small space and get much more output than trying to overdrive a lesser number of devices.

 

[Cemoi]

Luxeon Rebel Red is 40 lumens, Red-Orange is 50 lumens

Unfortunately the Star version is not available in red.
As a complete nube in the LED DIY world, I prefer an easy-to-solder star version.

 

[Oznog]

You can get a 3x Star board and pack reds on it. It's a solder board. However, the thermal resistance of that 3x board is not very good on a per-device basis. They seem to have only put the top copper foil in a small circle in the middle.

What do you actually want to do?

 

[LED Boatguy]

Unfortunately the Star version is not available in red.
As a complete nube in the LED DIY world, I prefer an easy-to-solder star version.

Stay tuned on that. Luxeon Star will have a line of blank Rebel MCPCBs in the next week or so. Figure maybe $1.69 each in small quantities. Grab up some stars, some loose rebels and solder paste, and learn how to reflow solder.

Hey Oznog: The 3 up Endor Star is something like 9 deg C/W/device. Do you suppose the boards from Asian Signals could be any worse? DANG those 100 boards for ~$35 looks tempting! Wonder if he'd fire me a couple samples of the 3 up square and 6 up round so I can do some testing.

Since I'm blabbing, does anyone have current-generation high-output greens or warm/neutral whites they want to unload?

 

[Cemoi]

What do you actually want to do?

DIY bike tail light.
Thanks for your replies.

 

[2xTrinity]

IMO for a bike taillight, any of those will likely be more than enough -- possibly too much. I had a Lux-III running at 350mA in a mag sized head with stippled reflector and frosted glass, and it was still MUCH more intense than typical car head lights, so much so that it would actually be creating glare in the eyes of drivers behind. What I eventually ended up doing with that was loading it up with a TIR optic from DX that projected the light into a horizontal line pattern, then aimed it at the ground, so that it would actually illuminate the white line separating the road from the bike lane, rather than aiming directly backward.

 

[Oznog]

Stay tuned on that. Luxeon Star will have a line of blank Rebel MCPCBs in the next week or so. Figure maybe $1.69 each in small quantities. Grab up some stars, some loose rebels and solder paste, and learn how to reflow solder.

Hey Oznog: The 3 up Endor Star is something like 9 deg C/W/device. Do you suppose the boards from Asian Signals could be any worse? DANG those 100 boards for ~$35 looks tempting! Wonder if he'd fire me a couple samples of the 3 up square and 6 up round so I can do some testing.

I got into some interesting debates and measurements on that here. I was proved wrong by my own measurements, the AS board performs well.

The 3-up Endor is 8.9C/W, by their own spec sheet. That's fairly poor because if you look at the right pic you can see the top copper spreader under the thermal pad is limited to a small circle in the middle.

The AS board measured- as best I could measure- at 6.93C/W per device when it's used 6-up. This is worse than any of the single-device MCPCBs but it's better than the 3x and considering that it's a 6x that's pretty good!

Oddly enough I did some calcs and showed that a normal 0.064" PCB with 1 oz copper can go down to ~4C/W per device anyways, with the right arrangement of thermal vias. I think MCPCB is unnecessary and counterproductive.

Just to be clear we're talking conduction to a rear heatsink. If the board is simply floated inside a housing then the dissipation is far far less and I can't estimate AS boards vs MCPCB. I think MCPCB may win due to much higher lateral thermal spreading, but it's still gonna be inadequate anyways for multi-watt configs.

For taillights, use red-orange (aka "Sunset Orange"). That is a wavelength that the human eye responds more strongly to that "red" and is fairly standard for taillights now.

 

[Cemoi]

What I eventually ended up doing with that was loading it up with a TIR optic from DX that projected the light into a horizontal line pattern

The L2 flare lens might do a good job giving a very wide beam. But I fear its bizarre shape will make it difficult to integrate in a DIY tail light project.

For taillights, use red-orange (aka "Sunset Orange").

OK, thanks for the advice.

 

[Oznog]

Well, just to be clear- the angles that the device throws and the point intensity that you can see are two different things.

For example, I could have an exposed, unfrosted lightbulb- visible filament- that casts light in 360 deg. That's potentially blinding. Place it in a frosted globe and it's soft and diffuse, same lumens, same angle.

Optics change the spread pattern, but this may or may not change the point intensity for the better. Car taillights use an ingenious series of prisms on the surface so that even though the light started from a point, it is redirected to light up the entire plane with an even glow so your eyes see the light being emitted from an area hundreds or thousands of times wider than the original point source (filament). Frosting "kinda" does this but not so well. You'd still see maybe a 1" lit circle on the glass if it's as close to the emitter as taillight glass is to the bulb.

 

[faklya]

I agree 2xTrinity, a red power led @350mA can be a little too much for a bike tailight. With around 40 lm of red light, it's light output would be around a typical car taillight, in a more compact size, that could disturb drivers. Like you were going braking all the way. But of course it depends a lot on the optics, on the angle of the beam.
I am also thinking on building a bike taillight myself, using a standard dynamo taillight. They usually have a 6V@100mA incandescent bulb, or 0.6W. If it's efficiency is around 10-15 lm/W, it could have originally about 6-9 lm of white incan. light. That filtered into red light it's probably below 5lm. So I think 10 lumens would be more than right for me, driving a red power led @100mA. But I want a brake light also. Here in europe I think standard car tail light is 5W and the brake 21W. So for the bike brake light I think I would need 4-5 times more lm. Driving @4-500mA should do that. At 500mA I need good thermal management though, but it wolud usually operate only some seconds that hard. Initially I wanted a red Rebel for this project, but considering that I have an ordinary reflector I think a side emitting led wolud be proper(Maybe a Lux III).
I'm not sure about the color tint, yes red-orange could be more visible, and more efficient, but personally I like darker tints. With red-orange it could look just like an ordinary taillight
What is your opinion about? What the beam pattern of a tail light should look like, where should it throw more light? Does worth trying a side emitting diode?
Thanks for the posts, very informative.

 

[LED Boatguy]

FYI Luxeonstar is now selling 3-up Rebel unpopulated MCPCBs along with some others.

http://www.luxeonstar.com/led-mounting-bases-c-53.php

Not much help to me as I just ordered the AsianSignals 6-up and 3-up boards. I see AS now has Cree MC-E boards too.

Too many choices

 

[Changchung]

What about K2??? I like those Leds...

http://www.luxeonstar.com/luxeon-k2-led-red-lambertian-75-lm-700ma-p-256.php

http://www.luxeonstar.com/luxeon-k2-star-red-75-lumens-700ma-p-131.php

 

[Cemoi]

a red power led @350mA can be a little too much for a bike tailight.

What about driving it directly off two NiMH in series?
According to e.g. the Luxeon I datasheet, the red-orange LED current @ Vf=2.7V (i.e. freshly charged batteries) is 220 mA.
Then once the batteries voltage stabilize around 2.4 - 2.5V the current will be between 120 and 150 mA, still quite bright I suppose?
The minimum Vf is 2.2V, which means a 50 mA current, still visible.
Batteries would be protected against overdischarge protection (1.1V per cell).

 

[znomit]

Three AA NiMH and a current limiting resistor will do a good job of keeping it somewhat regulated. Maybe a couple of resistors and a three way switch for high/low/off.
200mA will be very bright and run for 10hrs.

 

[nuggett]

If I may hijack a bit, I want to connect 2 red leds to augment my motorcycle brake lights, but to flash. Anyone have any idea for a flashing driver? 12 volts.

 

[znomit]

You can get a MC brake light flasher kit. Hook that to a 350mA buckpuck and drive the LEDs in series.

 

[Cemoi]

Three AA NiMH and a current limiting resistor will do a good job of keeping it somewhat regulated

Thanks for the suggestion, but I'd rather use two NiMH only. Will direct driving the LED (no resistor) be OK?