I am thinking that 3 x 7135s is enough for an led rated at 700 ma, even using a large host to take up the heat.
I am putting it in my UniqueFire UF-T20. Had to pry out the star, and melt out the driver. Need to file off 1 mm. from the deep red star. The driver fits fine.
What is the best way to deal with the heat the 7135s give off from the low forward voltage of the deep red led? I have used lifepo4 or 2 NiNH 18650s in other applications where lower voltage was helpful.
I used a 7135 based driver on a Oslon IR-led with very low Vf. I made some extra heatsinking for the 7135-chips and used a LiFePo battery, but the light also works ok with a normal Li-ion. It has become a bit long thread, sorry about that.
I have not yet managed that, but it is on my radar so to speak. I’m thinking I’ll buy some synthetic diamond powder and mix it with some silicone (maybe GE Silicone II) to use for potting.
I am pretty comfortable with this level of current (and subsequent heat) because the Rebels are on Al SinkPads. The heat transferring qualities are excellent on these, and I saw no detrimental effects to running the lights on high for an hour (other than a manageable level of heat)
But to be honest, I think the max drive currents on Rebels are extremely underrated (700mA-1A, depending on whether it’s an ES or regular Rebel). Both lights performed like champs when running on high for an hour, with no recognizable negative effects. I mean come on, the XP-G2’s maximum drive current is 1500mA, and it can do a little bit more than that
I tested a few other modern Luxeons (Q and Z) for output, voltage and current, and they go way over specs. I expect the newest Rebels have the exact same die as the Luxeon Z and Q, so the behaviour would be similar.
I have already finished it with 7135 x 3. The lower forward voltage increases the heat generated in the 7135s, but I read they can even handle 2 x Li ion. The driver board won’t get hot because it was designed to handle 7135 x 8, and it is soldered to the brass pill. So I think I am quite safe about heat.
My red XP-E looks somewhat orange next to it, but when I took a picture the Luxeon looks pink. It is brighter than I expected.
My DX lens is weak around the edges, but the effect is a nice halo around the spot.
So you think I should increase the drive current to 3A? Great, I do too
We’ll have to wait and see about that, but maybe I’ll fill the rest of the spots on the board and see what happens. If I don’t like it, I can always drop it down. Not difficult at all.
One thing that I’ve found confusing about the Luxeon Z Line of LEDs is that the pad layout. On the Luxeon Rebel and Luxeon Rebel ES, the pad layout is identical. However, the pad Layout for the Luxeon Z and Luxeon Z ES is entirely different—the “standard” Z doesn’t have a dedicated thermal pad. Just something weird I noticed when browsing Mouser, Phillips, and SinkPad’s website.
Hank, That’s sort of what I’m discussing with DJozz! The simple stuff first: you can use any 7135-based driver you would like to power these LEDs—like this one.
Now for the maximum drive current…that’s what we’re pondering. My first build ran these both at 1.4A (4x 7135-based driver), and they hardly broke a sweat. I have pushed them up to 2.1A, and they feel like they have a lot more to give. DJozz’s studies of the Luxeon Q and Luxeon Z suggest that they can perform at 3-4x their rated current ceiling and be quite happy doing so with the proper thermal precautions.
I think it is safe to say the 700mA is EXTREMELY conservative though, and I wouldn’t hesitate to go above & beyond that number.
It’s not super-special, it’s just a standard M1 pill, I chose the M1 b/c it works with 20mm stars. If that’s not helpful, I’ll certainly take pictures…just not quite sure what you’re looking for!
I’d like to see how you put it together as well as how the light looks. I plan on doing a similar light and maybe building a bigger light from an old 3D cell 21LED light that the LEDs no longer work on.
As you can see, the pill is pretty basic—screwed down emitter. I just used AS5 underneath. It’s a shame that this star isn’t offered in Cu (though demand certainly isn’t high enough for SinkPAD to justify the cost of tooling) so soldering the star to the pill wasn’t really an option. But I think a thin layer of AS5 on a lapped star that is tightly screwed to a brass pill is doing the job just fine.
I put kapton tape on the reflector because there isn’t an emitter spacer for the rebel.
I’m not equipped to reflow LEDs onto stars yet (maybe never) but have used Luxeon PC Rebels premade like this: http://www.luxeonstar.com/amber-10mm-rebel-square-coolbase-leds
They drop into the (rare) SK68s with solid pills just fine.
and the 20mm boards also have done fine in a generic 3xAAA light with a solid back to stick the board to.
Drilling and tapping for screws will probably be next for those.
They do get warm — and they’re astonishingly brighter than the old non-PC rebels.
Just learned that they will be available within a few weeks on SinkPAD II boards.
I’m waiting on those for my next experiments.
In email I mentioned and pointed to BLF and CPF and urged Luxeonstar to offer some kind of discount code, since their free shipping cutoff is high. Here’s hoping.
(By the way, the packing is astonishingly high end — mounted LEDs came stacked in little snapcap black plastic cases, labeled, and separated by rounds of antistatic foam. Impressive, since I’ve understood how semiconductors get slowly wounded, accumulating damage over time by static carelessness)
Seems that expressions of interest (and suggestions for improvement) are welcome:
I see that some flowers are deep red color, and that they are darker in white light than those that are medium red color.
With white light, these flowers are white at the tips of the petals and then red shading darker near the center.
With a Philips Luxeon deep red led, they are uniform light shaded.
Apparently the dye they use is dark only because our eyes are insensitive to deep red. Red colored artificial objects look equally light under the deep red light, but they look much brighter in white lite. I have noticed this with other flowers, but the white tips show how little this dark looking dye absorbs deep red light. Bees probably see this as a bright color, without the aid of a flashlight.
Another example is red wine. A glass of half wine and half water looks almost black under white light and almost clear with deep red.
I also have a lime led. Colors look sort of normal under it. There is red and blue but there is no purple. If there is also some white light, the lime led shadows look purple.
Flowers, so far, are the best objects to test color rendering.
Luxeonstar now offers the Rebels in colors on Sinkpad2 bases
and other variations
Canadian source, $100 required to get free shipping, so those real good at soldering tiny things will have better options (grin).
I bought several colors on their 10mm square bases — easily centered, on the few solid-pill AA lights I’ve been lottery-lucky to get.
If you contact them, urge a discount for flashlight modders; I’ve swapped emails with some folks there and pointed them to BLF/CPF, so maybe they will look in.