lol, good catch.
Thanks TA and Maukka! Finally some good 4000k 90+ CRI emitters. The FA3 is further below the BBL whereas the FD2 is still mostly above the BBL so I think the FA3 should have much better tint especially when used in mid and low output. It could be just the random sample though. Hope we can get a group buy for these soon to be extinct good tint SST-20.
Huh, guys I just noticed.
The SST-20 4000k is now available on Mouser.
Will ask for the tint bin.
Might be good like the 3000-3500k, but I really doubt it.
For some reason the WW (sub-3500k) and CW tint (5700k+) SST-20 are available below the BBL but the NW mid range tints are basically impossible to find. They have 100k FB4 tint bin LED’s ready to go though lol
All the 3000k on the market are good. However, didn’t the 3500k group buy you organized that Mouser said was GA3 tint bin unfortunately tested out to be above the BBL and likely a GB4 tint bin?
No, it was a GA3, right on the BBL. That’s what Mouser told me anyway.
The problem for these LEDs is that they tested very slightly below the BBL, or on the BBL.
That means while an LED might pass Luminus’ binning tests, it might not for us flashlights.
Actually it is slightly above BBL even without reflector or optics. Once you add the reflector or optics and/or run it on low output, it will always be above the BBL, so it is not like the 3000k HD2 tint bin that is always below the BBL. GA3 tint is supposed to be entirely below the BBL.
This is really useful data.
I was really hoping the FA3 bin would be below BBL even at low power, but it looks like it still isn’t. Great tint at higher power though. I should put some on a MCPCB to see how it looks during use.
Would it be safe to use SST-20 with a FET driver in 18650 light?
They are unobtainium apparently. TA keeping his stash secret for production lights 
A single SST-20 fries at about 7A, so triples or quads are your best bet when using them with a FET. Otherwise with a medium drain cell like NCR18650GA and a light that isn’t SUPER low resistance a single one should be fine, people have been using them in the D1S and other FET lights.
Texas_Ace had his supplier send me a few, totally out of the blue. I’m not sure why. I don’t have good test equipment like maukka does, and I’ve never actually reflowed LEDs onto a MCPCB. But perhaps it’s time to start.
Based on what I have seen in maukka’s review for CRI and output, the high CRI SST-20 is best used at around 1.5-3.5A as max current. Under 1.5A, tint is not pushed enough to or below the BBL, above 3A, but especially above 4A, color rendering for deep red (R9) starts to get somewhat worse (still above 60), and regarding efficiency, my choice is also max 3.5A, because when the current is raised above that, the gain in output is not really perceivable, but both heat production and battery runtimes go linearly worse - and above 5.5A, probably LED will degrade much faster during usage.
So if a FET is considered, a single emitter configuration will probably fry the emitter, unless other specific circumstances get the current down enough.
(But I would happily take a look at those posts, which confirm, that single emitter FET thrower builds are possible).
Thanks to those of us who actively attract regulated drivers there are now more of them. I recommend you to stop the anathema of FET drivers and start thinking higher. Believe right, there are no limits.
Well, I was worried that SST-20 and FET were not a great combo. But it was at the edge of my understanding of the stuff so I was not sure. Thank you all for the confirmation.
Yeah, that’s right.
FET+SST-20+low resistance circuit = poof. Magic smoke is released.
These newer SST-40/8A high current linears Simon requested being made feature an Atmel MCU onboard. Playing with the sense resistor also helps. I will use one as it comes for a White Flat 2mm² in a C8 but inside an S2+ with LH351D I will replace the stock R005 with an R010 (4A) or R010 plus R050 (4.8A).
They have solder mask down at the spring base. Can fix that with Cu sheet + thermal glue but it is another bummer for my spring bypass technique. 
:-)
Since your FET-driver crusade started, I never stopped thinking higher, and believe me, my thoughts have no limits. Unfortunately, reality does not listen to my wonderful thoughts and existing well-regulated drivers are often bulky or expensive, or at least have a medieval user interface that after I started using TomE’s and Toykeeper’s firmware, I will never tolerate.
So I keep thinking, and using my pleasant and convenient BLF-A6/bistro/biscotti/Narsil/Anduril flashlights, keep not noticing that regulation is not perfect, and in the meantime wait for the small and still powerful affordable current-regulated driver with TK-quality UI to appear from some heaven.
Same here. I love the idea of linear FET driven lights.
Power density is great.
Efficiency is slightly better than with our standard FET+1 or FET+N+1.
I don’t think the hardware is inherently expensive - though I am uncertain here.
Flexibility of adjusting current to match one’s needs is great. F.e. I’d love to PWM SST-20 with 2A at moon and limit to 4.5A at Turbo. Wouldn’t go particularly low but it would have good performance and consistently great light quality.
But such driver is not available.
There’s Texas Commander but it never took off.
There’s Lexel’s driver but it’s in progress.
There’s Led4Power driver but it lacks good UI.
There are several other drivers that have, yeah, medieval UI.
So for now we need to settle at sub-optimal choices. Like PWMing a FET to reduce current. Or adding large banks of 7135s. Or adding resistance. Or using only well-matched cells.
I really hate the state of affairs. But it is what it is and I am unable to do anything about it.
Reality does listen to your thoughts, albeit not everything you think is something you believe.
You allowed yourself to become addicted to those firmwares. Despite being your respectable choice, as any other addiction that is bad. And those last words I left in bold in the above quote are terrible.
Cheers :-)