7 Luxeon Rebel 40mm star - 2,100 lumens @ 1A - Maglight mod anyone?

I was checking out Luxeon's site (something I haven't done in years) and something caught my eye: a 7 x Cool White (5650K) Rebel 40mm round pcb that claims 2100 lumens @ only 1A!!! I know it wouldn't be the brightest light ever, but I sure love efficiency and that just seems great.... Seems like it would be an easy mod for a mag-modder (not my cup of tea). Has anyone considered or even tried this??

Shao

Hi, can you help me with a link ? I looked at the Philips Lumileds site and luxeon.com and could not find this product.

There are some other very interesting products there, but I didn't find the one you are describing.

Thanks

Harry

It looks like it would work, though it would be on par with an XP-G based 7up board. the Rebels run at 320 lumens, 2.85V, 1A, so 112lm/w. The Cree PCT says an XP-G R5 does 320 lumens at .9A, 3.35V, or 106Lm/W

Slight advantage to the rebel's not a ton, I'd call it 6 in one hand half a dozen in the other. (unless you want color! then definitely rebel all the way!)

Even with a 3x XM-L outputting 2100 lumens, (700 per LED) its at a 2A drive current, 106lm/w, still a wash, though slightly cheaper price for the LEDs.

a quad XM-L doing 2100 lumens (525 /LED ) is ~1.4A and 119lm/W. Slight advantage to the XM-L, again, not a ton, and youre spending the same on LEDs, with the only advantage being possibly bigger optics can do a tighter beam and throw further. (however I dont know of any quad XM-L optics, so again, wash)


7 of the rebels would be 19.95V divided by 16 cells is 1.25 V/cell so you could run it in a 16AA -> 4D adaptor with direct drive (maybe a resistor if they are too hot off the charger) You could also use a max-flex from taskled in a 3D mag package with 26650s



Certainly seems like a nice combination! When are you putting one together? :nana:

Hi, can you help me with a link ? I looked at the Philips Lumileds site and luxeon.com and could not find this product.

Click to expand...

The Luxeon website is http://www.luxeonstar.com

7 of the rebels would be 19.95V divided by 16 cells is 1.25 V/cell so you could run it in a 16AA -> 4D adaptor with direct drive

Click to expand...

According to their white papers the Rebel typically draws 2.85V at 350mA, and 3.1V at 1A, so the total would be 21.7V for 7 @ 1A, still within range for a 16AA setup though as that would be 1.35V per cell.

In theory couldn't you run them parallel to not require the 21.7V at 1A, instead only requiring the 3.1V but at 7A? I realize there are several problems with this though, wires everywhere for example, not to mention the short runtime. Could a 26650 even handle a 7A draw? I'm not an electrician, and I don't even play one on TV, but from what I understand the basic principle is when you run them in series the amps stay the same but you add up the voltage, and when you run them in parallel the voltage stays the same but you add up the amps. Am I a little off there?

Thanks for the link. They did a nice job of making up a lot of offerings.

I tend to use Rebels as my default LED for a variety of reasons, one being the ease of exactly defining (and buying) the CCT and CRI I need for a variety of projects. Other people like to use Cree as their default LED supplier - it all works.

The board has a very reasonable them path - they did a nice job of specifying it from the LED junction to the bottom of the PCB. A very nice implementation of the Philips Lumileds thermal data sheet. Make sure you have a good path for it to the light.

1 amp will work for a flashlight just fine, although at the top end of the spec. Not sure if you will gain much over a slightly lower drive current, but they are pretty tolerant with a good thermal path.

If you can run them in series, it will be better, as the Vf varies a bit LED to LED, regardless of brand.


Edit - Looking back at CKOD's post - he is dead on correct.

To run the 7 LEDs in series from a lower battery voltage, just use a "boost" driver, which takes in the power at a lower voltage, raises it, and regulates the current, as needed, for the LEDs. There is a long list of drivers in a thread in the battery / electronics section of the forums. It is under the "sticky section" at the top (threads of interest). example, taskled.com makes some nice ones.

The total power required stays the same (amps x volts into the driver = amps x volts out to the LEDs ) with of course some conversion loss.

If you are using a driver, one cell may or may not be able to supply that much current, but it is rare to find a driver that can boost that much power in one step and run at such a low input voltage. I don't know of any but it might be possible. 3 cells would be much better if you have room.

As far as 7 amps from once cell, a good rule of thumb for Li cells is to not run them at more than 2x their real capacity. Example, if a cell has a "real capacity" of 2,000 mah at the desired current load, then don't attempt to pull more than 4,000 ma (4 amps) from it. That assumes that you know the cell's "real" capacity at that load, not it's label capacity rating measured at some very low current draw.

Yes, I know lots of people that run them harder than that, but they have walked and read a lot before then ran a formula 500 race. They also are prepared if the engine throws a rod and the driver needs to be extracted from a fire.

Is that what people refer to as 1C or 2C, C being the "real" capacity of a cell? The problem I see with the 7 emitter board is that it would be very difficult to run, for example, 2 parallel sets of 3 emitters in series and then one all by itself. I was actually thinking it might be a fun setup to do all the outside emitters as white, and the middle one as red (or other color) and run them on different circuits/switches so that you can have a multi-colored flashlight.

One of the 4Ah IMR 26650s should keep up with a 7A load fine, but then the issue is making everything share current nicely. If your Vf binning, and thermal coupling between the LEDs good, then you might be able to make it work, otherwise youre looking at 3x 7135 chips for each LED ( 21 total) hope you have a lot of space for driver board. Though that would be one heck of a 1D mag. And while we're throwing thermal dissipation to the wind, you could do the same with a SST-90 also

But, you could have a microcontroller running all the 7135 chips and make for multicolor fun with a 7up board with each LED individually addressable.