This led gets my attention. I read through this thread and have more questions than answers.
My objective, in a nutshell, wearable on head worklight 400 lumen 4000 candela 12 deg Hotspot, 10 hours of cc, not green tint: I am interested in building more work headlamps that can get me 200 lpw for 10 hours, using 2s18650 and the buckpuck 700, 700 mA, with a dial dimmer. I use a 31mm,prefer 36mm aluminum smo, because I need a bare minimum 4000 lux at 1 meter. My last build was rain resistant, 2 series 18650 at rear of headband with buckpuck, the head pivoted (one build used thin copper pipe on aluminum with elastic) the led was an Cree XP-L v6 4000 Kelvin thermal epoxy mounted to a flat aluminum avid with finger spikes that epoxied to the aluminum reflector. This thermally epoxied to an old fashioned computer heatsink with thermally attached copper pipe for pivot. The open bezel design allows hyper cooling, with the aluminum reflector (thermally epoxied to pcb) as a huge part of the cooling. This allows a very low led junction temperature. I recall under 40C using mm probe and ir. At 700 mA I measured 373 lumens, over 4000 lux, which I could sustain 9.5 hours. My last build like these was in 2016. I have seen led technology marching backwards, chasing lumens and allowing spectrum filters for higher cri…… I might be interested in 80 cri, but I do not do photography, I only need to see subtle color differences, which even 65 cri metal halides do excellently—(cri+gai+luminance equals color, not just cri). I did attempt to build some 2x amc7135 lights but I really felt the drop in efficiency from 92% to 70% made the lights useless for my jobsite purpose. I really can’t do anything with a runtime much less than 10 hours.
From reading this thread, it looks like this led might bump up the runtime, while allowing the same throw/lux in a 31 mm at 700 mA. However, I am reading that the color will be unacceptable green at 300 to 700 mA. I am thinking that the sbt 70 would offer poor lux at the max allowed mah for a wearable worklight, which is 700 mA. (yes a 3s 1850, or 2s 21700, would allow 1 amp for near 10 hours, however, weight, charge time, and number of bays for 3 to 6 workers, gets messy).
The XP-L2 has lower forward voltage, but it has lower output at 700 mA, which is a big deal since 31 mm is available, while 36 is not. I do not like 40 mm Hotspot size. And at 31mm, my XP-L v6 barely got 4000 lux (minimum acceptable 2meter 1000 lux, for inspection). So, I am thinking the tradeoff of runtime for lower lux with 700 mA, not worth going that route.
I am also thinking that this led might not cut it either. (seriously, no advancement in leds in 4 years, other than wow factor and filtered spectrum to better cri, lower output? However, if output same, lower fV, longer runtime is a great plus since I am still struggling to get 4000 candela for 10 hours. But I hate green tint. I am happy with the 4000 Kelvin XP-L V6 70 cri, tint-although I would like to bump up to 80 cri, if I could without significant output loss. … I have purchased 3 95 cri lights in the last 3 years, including on fixed metal halide, but apart from the tint novelt, I am not really happy with loss in output tradeoff. I use them more as reference lights for my main work lights. )
Now, my un edited questions and notes that arose from reading this long thread, if anyone cares to read and answer:
1. What is the binning? (Flux : N4, Color : BA, Voltage: VK?)
2. Lpw at 1 or 2 watts? At 350 ma and 700 ma?
3. Gallery comparing color rendering v. Xpl v6 at same color temperature.
4. I take it that the fV is 15% lower than the xp-l v6. People are reporting it runs at same due heat as the xp-l v6 at same current. So, I am guessing a 15% longer runtime until the 7135 linear driver cuts out. This correct? Then, what about a 700 ma buckpuck, 92% efficienct buck driver v. A 70% efficient linear driver?
5. Having butchered every led I attempted to dedomed with razor and gasoline, I would rather buy a predomed sbt-40 that has been certified with a bin rating, above the xp-l v5.
6. I don’t understand why the sbt-90.2 is not smaller than the xp-l, 3030 v 3535, at a glance on spec sheet.
7. I don’t understand how the second generation of the sbt-90.2 fits in with the sbt-40: how much lower is the lux per watt as compared to the sbt-40 ; why am I not seeing the sbt-40 having a second generation like the sbt-90.2?
8. Are there any 2s18650 buck driver lights with ramping, which are headlamps?
9. Apart from the buckpuck with dimmer, are there any affordable buck drivers out there that target useful full workday runtime off 2 cells?
10. Do the dome defects and inconsistent suppliers make the led not worth ditching Cree? Is baking at a low temp for a while able to drive out humidity and avoid reflow bubbles?
11. Kaidomain.com, banggood, intl-outdoor, digikey? Aliexpress safe seller , RMM and IOS?, arrow? Mouser, mtnelectronis. Shenzhen Great Technology Co.,Ltd.?
12. Reflow onto same star as xpl? Lead solder paste Best, still in 2020?
13. Ring artifacts?
14. Any fixed lighting using the led?
15. What washer (thickness) could be used to dedomed, by protecting the die? Single edge, razor, scalpel, or beard razor, would work best? Detail Scissors? And, if I don’t get all of the dome off does it ruin the throw? Is it a really thin line between nicking the die and getting a dedomed led that is an improvement?… Toluene better than gas, xylene, lacquer thinner, acetone? (toluene is 5 minutes. Vinh stopped fully removing SST-40 domes, now he shaves them. He says it gives him the same throw, but better output.)
16. Best place to buy 36 mm or 31 mm smo reflectors? 26 mm, maybe for a dedomed version?
17. If I reflow myself, how to avoid creating bubbles, as is the case with most of these?
18. Tint lottery?
19. How badly does the tint shift during amp/mA change? It would suck if green if not slamming with pwm to highest current!! (twisted raven wrote: My SST-40 gives me a brilliant cool white with little green at max output, but at lower outputs on a non PWM driver, it gets rather green. My Quark 2AA-X does that. On turbo, it’s not too bad. On moonlight, greeeeeeeeeen. )
20. Voltage bin? VH, 2.5 VOLT, only good bin to beat the xp-l? (vj 2.7v, vk 2.9v,vl 3.1)
21. If a fet is a dd, with no resistor as mnt electronics claim, why would I pay $20 for a wiring bypass. Mnt can not be correct in what a fet driver is.
22. This is lower than the Cree v6 xpl version 1: Nw only 310 lm typical at 700 mA
170 lm/W typical at 700 ma. Why so much lower? VH and n5 bin in only 6500k?
23. Is this true?: “the die is larger than XP-L or XM-L2, so you either need large optics or a lot of power (or both) to get impressive throw numbers.” (The die is maybe slightly larger. I measured about 10% less throw than dedomed XPL V6 at the same current).
24. Has luxeon or Cree improved in the last 3 years. How about a xp-l v6 in 80 cri? (green tint in a useful =<.1C drive level makes the led useless, other than search and rescue and ocd pursuit).
Hank noctigons from hank, kd, VK from Ali, reflowed on noctigon, Kaidomain.com KD DTP copper MCPCB-s., or aluminum
21.04€ |10 stücke 100 PCS Luminus SST 40 10 watt LED 1 100 lm Kühles Weiß 7000 karat statt CREE XML2 XM L2 LED Licht Emitter Diode für taschenlampe|cool white|10w ledled 10w - AliExpress
https://www.aliexpress.com/item/10PCS-100PCS-Luminus-SST-40-10W-LED-1100lm-Cool-White-7000K-instead-of-CREE-XML2-XM/32819284669.html
Funny, it appears renamed to “GreatLED Store” from shenzen
Like KDlight DTP Mcpcb is 1.5mm thick, Noctigon is 1.6mm thick and SinkPAD-ii is 1.65mm thick. (Is there anymore available DTP Mcpcb’s out there?)
( Vinh uses VTC5A)
KD has VH Voltage binned SST-40’s now.
Has higher brightness bin too.
http://kaidomain.com/Flashlight-DIY-and-Tools/LED-Emitters-or-LED-Star-on-PCB/High-Power-Luminus-SST-40-Voltage-VH-N5-BC-White-6500K-LED-Emitter?sort=p.date_available&order=DESC
Nw only 310 lm typical at 700 mA
170 lm/W typical at 700 mA
https://www.mouser.ca/ProductDetail/Luminus-Devices/SST-40-WDS-F50-N3501
https://www.mouser.com/Search/Refine?Keyword=sst-40