Epic!
I obviously don’t have a light that pushes an XHP70.2 at 95W, but my K30-GT with its 57mm head easily out-throws my L6 with 75mm head. The SBT-90 is interesting because it can hit similar output levels in a realistic build and achieve similar or better intensity with a much smaller head. Now we can have soda can lights with a single emitter that can reach over 1000m.
If the 90.2 is like the old 90 then it has a piece of glass in front of the die which needs to be removed if you want good performance using a collar.
The CFT is bare die like the black/white flat osram LEDs.
Yes, it has a piece of glass in the front.
I have used quite a few circular die SBT-70 as it’s one of my favorite emitters, that window over the die has been removed with rare exception. I have killed a $70 emitter be ause of it, slipped with my soldering iron and trashed bond wires. But I still remove the window because it works better.
There are 4 Gen 2 90’s on the way here for me to play with, and I will see how this one stacks up against a CFT-90. I have a 124mm reflector that’s in a light I built from scratch which seems the most likely candidate, and I have an old light that seems to be needing attention (Crelant 7G5 with a Tofty switch, even considering marrying he 7G5 head to a D size Maglite tube for the larger battery) that will also likely become a host for this big contender. We’ll see how it goes.
There seems to be a lot of variability in the output measurements for this LED. I assumed it was mostly because of calibration differences of different users methods, as is usual for trying to accurately measure lumens. But your comment made me wonder whether maybe some of the variability was heat path affecting the output vs current curve of the LED. The only current dependent measurement I have seen is the one linked in the first post of this thread (for the closely related CFT90). It shows the output increasing still at 20A, but definitely with a decreased slope.
For what it’s worth, I have measured the relative output of the SBT90.2 in my convoy L2. I reflowed the LED on a 20mm MTG DTP board. I connected my 20A max power supply to the light and measured the relative output by ceiling bounce. I started at low current and did the measurements quickly, pausing only 4-5s at each plotted current point. The light was warm at the end of the test.
The plot doesn’t tell us a whole lot precisely. It does tell me there isn’t something catastrophically wrong with my heat path, since the curve looks pretty similar to the CFT90 test.
I don’t think you were reading closely enough.
Kawiboy says the factory specs look accurate. The factory measures raw leds with a power supply. Kawiboy measured OTF (out the front) of a flashlight using batteries. When your measuring leds in a flashlight you get losses from the reflector, lens, driver and voltage sag from the battery. The reflector is the biggest loss and might reduce output by 20% or more. This explains Kawiboys lower output levels.
So there is NOT a lot of variability in the output measurements.
I’m well aware of optical losses responsible for OTF vs LED lumens, so there is no misunderstanding. I was referring to reported OTF measurements that I’ve seen around the forum. There is always some variation in reported lumen output from different calibration, but perhaps it’s not any more than usual.
Okay, yes, there can be some variation in the way it’s measured. I believe Kawiboy has a nice setup to measure lumens and maybe a Maukka calibration light. Other people, who knows. They might use a shoe box and cell phone camera app to measure lumens. When you see someone state their measured lumens, find out how they measured it. If they use a questionable method, ignore their results.
I’m saying the LED output is right inline with the data sheet… @18-19amps…even with the losses…
5020lm at T/O…… 4315lm at 30sec……then checked T/O again after the 30sec run at 4590lm
. Now I put the same LED and 3 other LED’s I got from Wendy at even higher current readings and didn’t see anything over 5200lms at turn on, when driven past 20amps up to 25-34amps… How?.. ( I don’t have a fancy power supply), I can put up to 4 30T batteries in parallel as a power supply, and I didn’t see 6000-6500-7000lm as stated by some manufactures and sellers of this emitter in different lights being driven 27-30 amps and neither is anyone who owns these lights…from what is being reported by these owners with Lumen tubes…
. I think these SBT-90.2 are being seriously over driven IMO from what I have seen in real life… but maybe the Higher Binned SBT-90.2’s might hit higher, just not the examples/samples I have…
We all know flashlight manufacturers can make exaggerated claims.
Of course, leds tend to have a hill shaped output curve. As the amps get really high the heat increases which reduces output. You can keep pumping in current, assuming the led does not burn up, and output will start to drop. So you might be at a temperature limit or it could be you have some lower bin leds that simply can’t put out any more. IDK.
Aw come on Kawi, there’s gotta be a little more in there somewhere… 
.
That is why i have not purchased one yet, ” the Higher Binned SBT-90.2’s might hit higher “, and the data sheet shows they will. The problem is they are not for sale anywhere I can find them. Is the Military gobbling them up ?
Maybe?…. meanwhile… in a castle on a mountain top… 
Luminus hasn’t answered any of the half dozen or so e-mails I have sent, Mouser and Digi-Key are clue-less… :person_facepalming: so far…
The higher-binned parts might not even exist. It could just be a theoretical bin for a future part that they anticipate will be available when yields increase. Not uncommon in the semiconductor industry.
True…… you never know unless you ask, show some interest…give examples of interest. Not that enthusiast or flashlight makers are a major deal to them…. 
Well, I guess I will find out tomorrow as to how these behave. Probably should have picked up a set of 25S’s before now…
Thanks for this test, the first serious test of this led, what it is worth is that it tells a lot that a modder needs to know to use this led in a build: this emitter can do 20A without maxing out, and judged from the curve 20A seems a good current to aim for because it is a bit under the max but seems not too far from it.