If you add resistance with a fet driver you can set the best current for only one voltage level. If you set resistance with full battery when battery depletes the current will fall and output also. If you put in a slightly better battery it will give more amps to led. If you put in a GA cell it will give lower current. But a buck driver can make constant current for the led and no matter what battery you put in the led can’t get too much current, and if voltage dropping it can get more amps from battery to keep the same current at led. Simply it can keep more stable output.
Is there an option to get an 5000k 70CRI SST-20 in the Emisar D1S?
Having a 2mm2 sized die, it would give us quite a bit more throw coming stock, and would easily handle a FET driver.
We could even dedome it if we wish for even higher throw.
But so can a linear CC driver.
Sure, a buck driver is very nice and efficient, but it also has a larger minimum voltage drop than a linear CC driver like the ones from led4power or the LD25.
A buck driver can convert extra voltage to current. A linear driver burns it off as heat. With a very low Vf emitter, that heat can become excessive.
I do believe I saw someone custom ordered a D1S with an SST-20, but I don’t think it has been delivered yet.
My Astrolux SS with SST-20 pulls 5.2A through my ammeter with a VTC6 and 16AWG 8” leads. A D1S would probably pull a little more, as the Emisar lights take more steps to reduce resistance, but it probably wouldn’t hit the 7A known to be fatal to these emitters. I used the 4000K, 95 CRI SST-20 and still got a modest increase in throw due to the extra current. In a regulated light where I replaced an XP-L HI with the same SST-20, throw stayed exactly the same.
So yes, I expect the low-CRI SST-20 would work in the D1S with the stock driver and probably increase throw substantially. It might be best to use medium-drain batteries if you don’t have spare emitters and the ability to change them until we have more collective experience with this.
really? show us some photos…
Damn YMMV… it’s actually a black flat but anyway they’re not sturdier.
Huge difference on the forward voltage between black flat and CSLNM1.TG.
Black flat can be run on DD driver no problem, white flat pulls over 5A with Panasonic NCR PF, so 30Q will probably get it over 7A and that its a lot of current for that small led. Some guys here do not recommend over 4.5A for maximum performance.
Maximum output of the little White Flat is @ 5 Amperes.
But it’s not much more output then on 4.5 Amperes.
Not worth overdriving it that hard.
You probably won’t even notice the difference between 4 Amperes and 5 Amperes.
will the gasket work with GT mini reflector / L4P 3030 PCB ?
would it help if I design a CC FET driver for the D1S that simply act as variable resistor limiting the current,
I found a method that can be used to enhance thermal properties of the board, or we make a board like LED4Power has with fet on MCPCB
similar to TA or LD-4B or that cancelled BLF lantern CC driver I almost had final from design
Yes please. I know I don’t have a say. But I really want this light to become a reality.
Personally I am not concerned about the low end, low end efficiency, etc. on a light with this LED. The only issue here I see is regulating the max output. I don't understand the differences of your two options, the pros/cons, etc. Again, my own opinion, but I don't think it's important to have smooth ramping - a stepped ramping (mode levels) like what Anduril does would be fine, if that makes it any easier. Again, no idea what the costs/design tradeoff's of your two options are, so hard to say. I "think" an L4P design still has thermal issues - would the CC FET not have those thermal issues?
Of course it would be great if output doesn't drop as the cell drops, however that may be achieved. Perhaps the firmware can read the voltage and lower the resistance accordingly? I think this was discussed somewhere...
thermal issues with about 3W on the board no
I plan to reflow additional copper pieces on both sides to spread heat, bottom would conduct directly in the tube
but best would be a combined MCPCB with FET like L4P has
Don't recall what the D1S looks like inside, but Hank's D4's are extremely tight for driver height - ugh, but you are probably aware of that already. I'm not a fan - I'd rather add 1-2 mm to the pill size and overall length. More slop space, little more metal in critical heat area.
I am talking only about 0.5mm copper sheet or so, that would minimal increase the gap between tube rim and head,
but could be simply eliminated by shortening the tube on lathe, but MCPCB variant is best
Lexel, such driver would be great for many lights even if it required buying MCPCBs from Led4Power.
That because:
- Led4Power drivers don’t support e-switch (though that’s likely to change rather sooner than later)
- Led4Power drivers don’t allow the great open source UIs that BLFers made (and that’s unlikely to change).
Received Flat White LED with centre ring so I put it together with D1S stock driver. Center ring helps but it doesn’t fully focus the beam (i.e. hotspot is not perfectly round).
Becareful with center ring as the LED silicon is fragile. Mine got chip at the corner but still working ok 
Reflector size (Head to head): D1S have slighly larger reflector so more advantage in throw
Hot spot comparison from left to right
Convoy Flat Black, D1S White Flat , Convoy XP-L HI
Lux reading for hotspot at 8m distance
Convoy Flat Black = 3,392 lux
D1S White Flat = 2,897 lux
Convoy XPL-HI = 1690 lux
Clearly XPL- HI is no match with Osram LED
It’s strange that Flat Black out perform D1S white Flat . Maybe i need to focus the beam again on weekend.
simple answer is that its getting too much current and dropping in brightness again because of Auger recombination
White flat should run at 4-4.5A in a light not more
I am sure it will blast the C8 at 1A where the AMCs are fully regulating on both lights
What cell did you use in the D1S?