Of course, THIS thread is about the CFT-90 which outperforms the 90.2.
I know this is really expensive but how much better is it than the SBT90.2?
The CFT can handle a lot (a LOT) more amps, pretty easily makes a thousand+ more lumens.
It doesn’t look like that to me.
CFT-90 maxes out doing 5750 lm at 45A and does 5000 lm at 25A:
SBT 90.2 does 5000 lm OTF in a host, at ~27-28A which is about 5500 LED lm.
From these numbers, SBT90.2 would actually be better.
There is quite a lot of variability in these tests and I keep that as uncertain until I see further data but it doesn’t seem like CFT-90 is so much better.
I have the CFT in a light running off a single 32650 at 19A making right at 5000 lumens out the front. FET+1 driver.
My memory biting me in the butt again…
I looked back through notes and messages to a friend and after I put the Tofty switch in my 95mm scratch build with the CFT-90 it’s doing 5010 lumens at over 25.8A tail measurement. A 5M test showed 1.13Mcd for 2126.03M (1.32 miles). This was Sept 16, 2019.
Again it’s an FET+1 driver with Bistro that I built for it, running off a single Sanyo 30T 21700 cell. I DO use a 32650 in this light, but I tested it with the 21700 for max results to get an idea of it’s potential.
Edit: A pic of the CFT-90 in DBC-05 blasting fog just before sunrise…
So…within error margin of that SBT90.2.
The Luminus data sheets show the CFT much more efficient, probably because of the thick copper base it’s built on. At any rate, the CFT makes more output lumens at the same amperage and of course at the time I got the CFT the 90.2 wasn’t available to buy. I am hearing more about the 90.2 now and a few folks are having them specially shipped at a fairly substantial cost but I still haven’t located any for purchase. Probably won’t be able to resist if/when I find some. 
Unfortunately I can’t do a comparative test as the light my CFT is in is a one-off with a prototype reflector that was not put into production so I can’t duplicate the light build. Still, it will be interesting and I do have an Sup-Beam K50 that would house it nicely and has a really good reflector if not quite so large. I may also consider modifying my TN-42 for a 90.2 if I can find one.
Edit: The new Noctigon K1 is on the way to me and should be arriving soon, while it seems outstanding with the 1x1 White Flat it may be a great host for the 90.2, remains to be seen. 
SBT-90.2 is not present on Luminus website so I guess you’re referring to that 2016 SBT-90 data? Yes, that emitter was very inefficient…
CFT-90 appeared a couple of years ago. It has superb thermals but is also quite inefficient. SBT90.2 is younger by about 2 years. It is possible that they have a new, better die that wins with CFT-90 despite having inferior thermals.
But maybe they have the same die and CFT-90 is indeed better but different folks measuring different LEDs cause confusion.
Or maybe the truth is yet different. I don’t know.
Good points all.
The 90.2 in a light has been quoted at 5000 lumens with 27-28A power draw. I know my CFT-90 does that at 2A less. Unknown to me if the difference is in the driver or the emitter, reflector or lens. Or some of all those things. I do have a UCLp lens over mine.
Time will tell, I’m sure. And when I find one that I can buy it is very likely I will have the 90.2 in something… just a matter of time.
Best damn point for not getting Luminous chip, almost shelled out to buy one.
At what voltage 70.2 are being driven at 8A?
If it’s the 12v configuration then these parts are more or less equal in terms of performance, with 70.2 edging out in terms of being more widely available not needing as big wires and 90.2 for being able to run at 3-ish volt (so you can use many drivers in a pinch). 90.2 would performs marginally better still (that would mean 70.2 are to be driven at 32 amps in 3V setup to achieve that).
If the 70.2 in question is at 6v config then… yikes. 90.2 would then be severely outmatched.
Another point to stay away for 90.2 is the fact that you need a huge wire to push 30A through it. I don’t think it’s fun to run a 10-guage wire inside the flashlight.
Which led do you mean? CFT90 or SBT90.2?
The XHP70.2 is a powerful emitter, but it’s really in a separate class from the CFT90 or SBT90.2. The XHP70.2 has nearly double the light emitting surface area which is part of the reason it’s more efficient at a given lumen output. The CFT90/SBT90.2 is interesting because of its high current density and high luminance, and lower efficiency is just the price we pay for overdriving the LED.
It's very likely that the SBT90.2 uses the same die as the CFT-90. It's just a different form factor with slightly (!) less good thermal performance. There should be a difference at very, very high currents (this is only theoretical though because it's almost impossible to cool these LEDs at 40A+ inside a flashlight). I think it's great that we can now get these LEDs for 1/3 the price. They beat every other LED on the market if you want a high-output thrower with a single LED.
Specially over driving the SBT-90.2 when the data sheet for the SBT-90 gen 2 states up to 5400lm at 18amps…and were seeing it being driven at 25-30amps and claims of 6000-6500lm… the 3 I had barely broke 5000lm/22.5amps at turn on…
Large, monolithic chip with uniform emitting area of 9mm2 Extremely high optical output: up to 5,400 lumens at 18A from a single chip. Un-encapsulated die ideal for optical coupling in Etenduelimited applications.
5700K typical color temperature. High thermal conductivity package - junction to case thermal resistance of only 0.5 °C/W. Compact 10 x 11mm2 isolated surface mount package. Environmentally friendly, compliant with RoHS and REACH requirements
You measured bare led with power supply like the factory did?
Nope. I believe the factory rating, just from what I have seen 18-20amps at the LED seems to be the Max Output for the 3 samples I tested… (in actual lights)… 25-30amps is extremely over driving this LED…is the point I was trying to make…which some have tested in new lights….
.
OK, I can cool it, not a problem. But is it worth the effort and will it out throw a 70.2 at 7.9amps 12v. ?
If so, where can i purchase one on a copper mcpcb and a driver to push it to max output amps ??
.
.
Epic!