i think we are out topic now, we shold go back main topic
so
there is SST-90 led, with use single cell, and produces lots of lumen
the biggest drawback is high price
I expect if there is an XML3 that it will be similar to the XPL2 in die size and output.
The SST90 is interesting; we don’t see many incorporated in mods. It should be able to pull around 15A from a good single cell and provide ~3500 lumens I estimate, based on data from the datasheet.
If you are interested in high output 3V emitters, keep your eye on the Nichia E21A and E17A. They are 3V LEDs with small packages that can be grouped together to make XHP70-like configurations. There are currently heat dissipation issues because of the non-standard footprint of the emitters, however. Nichia E17A/E21A series (April 19th: updated with output tests in the OP)
Am I the only one not impressed by Cree new leds for flashlights?
I mean, tint shift, not easy to dedome, for our goal tricky VF, come on, people scouring LEDs of 1 generation before.
I get flashlight is not main focus so sure the overall idea of Cree makes sense to them, but for us so idk?
Seeing the XHP50 cross get smaller is nice true.
XML2 is a huge footprint for Cree, more material, less units made in one go, they practically scream that this huge amount of space is for different LEDs, not single die, the cost of making a XPL sized led doing a respectable lumens per watt is so much lower, I’d be surprised if the XML line would not be updated and vanishes when the machines making them cannot do it reliable.
Could be wrong of course.
I have nothing to prove. It is only my opinion. I have my own criteria. For example driver should drive modern powerful led xhp/ 144A with 1 *18650 and should has good stabilisation.
As xml3 is still in our dreams and has chance to stay there.(I think xml3 will not be in production because of hi current. 2000lm it is more than 6A.)
Are you doing driver development?
If so do you mind me pmming you for drivers are beyond me, but I am involved in some BLF projects drivers play a critical role in 
No:) I simple use several of above derivers in my mods.If you need I may address your questions to developers on fonarevka.ru.
Seems to be a lot of cool stuff going on at fonarevka.ru, some serious driver development. But have you tried google translate on technical discussions? Most of what’s said over there useless to most of us here.
Thanks, luckily for me our own DEL and Tom are two solid rocks, besides all those really skilled people like before poster Mike C
Could you ask on your forum if people speaking English to get involved here? Would be awesome!
Mike, I would not call this “technical discussions”, rather flooding. Usefull things like curcuit shematics, layouts and firware does not require even google translate, just trust that most other words cant help you.
so, i guess because the huge footprint, cree refer making xp-l2 led more
but, i dont want a dome with 4 side chopped, just my opinion
Isn’t Cree still making XM-l2 led, ii guess they can reuse their old machine, change something(material) and ta da, xm-l3
Better stock up on your favorite LEDs while they’re still available.
XM-L3?
You can already buy it in Poland:
Link
Thousands of lumens, thousands of mAh, kilometers of range, mysterious military technology, and zoomie of course.
And there is no way to stop this unfair sellers. Sadly, unconscious people are buying a lot of this crap.
sure there is.but its more complex and takes more space to make.
thats how the altilon 4s modules are driven in an automotive application.they are headlight modules.
Here I thought I was the only one to realize this since the XP-G3? 
:+1:
But I look forward to the better street lights as a consequence! :person_facepalming:
Bout the drivers, a LED driver is no rocket science. Not saying it's not time consuming, requiring some research and testing, but it's not crazy complicated either, least for an EE designer. In some cases, you just copy a reference design from the chip manufacturer. The problem with the better compact driver designs is either not enough power, too expensive, not suited well for a board layout that OSHPark can do, or they are not open source. The LD class drivers are all commercial, and the GXB20 looks like it's going commercial, so I have little interest in them, as compared to fully open source drivers.
Technically of course, the LD-2 family of drivers is superior in many way from our classic PWM FET/7135 designs, but high amps has some issues, and lacking being open source will always be a limitation here. So it could have, should have buried the classic drivers, but I don't see it happening in the foreseeable future. led4power has done a great job though with this product line, and offering a rich set of diverse options based on feedback here.
For the driver that eventually got used in the Meteor, not sure what happened there, but not having a presence on a English speaking site is a big disadvantage, and also I thought the driver firmware was at least partially written in assembler with comments in Russian - not a good combo.
Buck drivers in general are big and expensive, so will always be in somewhat limited use. Fortunately our PWM based drivers do handle 6V and 12V LED's with limitations, but not limited in power.
If XML3's do happen, which I don't really see or understand the need for them (XPL2's and XHP50.2's kind of fill the void), they should have the efficiency advantage via the lower Vf, and hopefully can take higher amps than the current XML2's do. In a way I'd rather see higher bins in the XML2, or at least some NW tints in the U4 bin.
Buck-Boost is certainly possible, it requires 2 switches and a single inductor. The biggest challenge is making it flashlight sized. I am in the very early stages of trying to design a Buck-Boost-DD driver on my own. I will make a thread about it after i try out a proof of concept with some components.
A lot of the driver veterans may not like it, MSP430, and very tiny FET packages. Also probably fairly expensive compared to DD or 7135 solutions, but hopefully more efficient. Ideally I’d like to be able to run any LED off of any combination of 1-4 lithium cells. We will see how much control I actually get though.