S21A LMP LML2AW.DC 3000K 100%mode
3 Thanks
| Part No. | Production | Spec. | Size | Color | IF | VF | CRI | Luminous | Radiant | Luminous | Directivity | ||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Status | L×W×H | (mA) | Typ(V) | Flux | Flux | Intensity | (degree) | ||||||
| (mm) | Ra | R9 | Typ(lm) | Typ(mW) | Typ(cd) | ||||||||
| Min | Min | ||||||||||||
| NF2L757GR-F1 | 3×3×0.65 | Warm White | 65 | 2.76 | 95 | 80 | 24.1 | 120 | |||||
| NF2L757G-F1 | |||||||||||||
| (Optisolis) | |||||||||||||
| NF2W757GR-F1 | 3×3×0.65 | White | 65 | 2.9 | 95 | 85 | 23 | 120 | |||||
| NF2W757G-F1 | |||||||||||||
| (Optisolis) | |||||||||||||
| NTCWS024B | 15×12×2 | Warm White | 135 | 34.9 | 95 | 80 | 515 | 120 | |||||
| NTCWS024B-V3 | |||||||||||||
| (Optisolis) | |||||||||||||
| NTCWS024B | 15×12×2 | White | 120 | 35.8 | 94 | 70 | 455 | 120 | |||||
| NFCWS024B-V3 | |||||||||||||
| (Optisolis) | |||||||||||||
| NFCWL036B-V1 | 19×16×2 | White | 180 | 35.8 | 94 | 70 | 710 | 120 | |||||
| NFCWL036B-V3 | |||||||||||||
| (Optisolis, White) | |||||||||||||
| NFCWL036B-V1 | 19×16×2 | Warm White | 260 | 34.6 | 95 | 80 | 1030 | 120 | |||||
| NFCWL036B-V3 | |||||||||||||
| (Optisolis, Warm White) | |||||||||||||
| NFCWL060B-V1 | 19×16×2 | White | 300 | 35.8 | 94 | 70 | 1200 | 120 | |||||
| NFCWL060B-V3 | |||||||||||||
| (Optisolis, White) | |||||||||||||
| NFCWL060B-V1 | 19×16×2 | Warm White | 460 | 34.8 | 95 | 80 | 120 | ||||||
| NFCWL060B-V3 | |||||||||||||
| (Optisolis, Warm White) | |||||||||||||
| NFCWJ108B-V1 | 24×19×2 | White | 540 | 35.8 | 94 | 70 | 2150 | 120 | |||||
| NFCWJ108B-V3 | |||||||||||||
| (Optisolis, White) | |||||||||||||
| NFCWJ108B-V1 | 24×19×2 | Warm White | 860 | 34.9 | 95 | 80 | 3380 | ||||||
| NFCWJ108B-V3 | |||||||||||||
| (Optisolis, Warm White) |
This is the result of my search. Some of them look too small, and some of them look too big.
| Part No. | Color | Size[mm] | LES[mm] | Power[W] | Vf[V] | Flux[lm, @ Tj=25, 85℃] | If[mA] | CCT[K] | CRI | Certificate |
|---|---|---|---|---|---|---|---|---|---|---|
| [SunLike] SAWS226AA | 28.0X28.0 | 22 | 38 | 35.4 | 4200 ｜3795 | 1050 | 3000 | 97 | 6000h Test | |
| [SunLike] SAWS1566A | 19.0X19.0 | 14.5 | 25 | 35.2 | 2741 ｜2467 | 720 | 3000 | 97 | LM80 | |
| [SunLike] SAWS1564A | 19.0X19.0 | 14.5 | 15 | 35.0 | 1749 ｜1574 | 430 | 3000 | 97 | LM80 | |
| [SunLike] SAWS1063A | 13.5X13.5 | 9.8 | 10 | 35.8 | 1094 ｜985 | 290 | 3000 | 97 | LM80 | |
| [SunLike] SAWS0661A | 13.5X13.5 | 6.0 | 6 | 37.1 | 614 ｜553 | 170 | 3000 | 97 | LM80 |
Are these voltages too high?
Yeah, I know that.
But the supply is unstable.
These seem great but omg that voltage
Simon, thank you so much for looking into this! Sorry I did not give more info on exactly what LED. I’ll link some BLF tests of them and try to update this comment with the exact part number or official website link.
There are many variants of the Seoul Sunlike, even in the 3030 class. One of them is 6V, tested here by maukka, official website here. There are many 3V variants, one of which is tested here by djozz; the actual part is unspecified but possibly this or this. Out of all the 3030 variants, the STWSC12S-E1H10000 can handle much more power compared to the rest and should be the one to get.
Here is a test by koef3 of the Nichia Optisolis; the main interest is the 5000K variant, which uses a violet die; official website here.
These are the same methods that Lumintop used. This looks like a safe way. But converting is changing a post and ring method over to this, not sure how. Probably should be built from the ground up as a dedicated single cell light.
It’s 4.2 volts insulated with a thick plastic donut. Lumintop permanently glues their first gen retaining ring. I’ve accidentally shorted an 18650, not going to do that by a factor of 10+ if I can help it. Just for safety sake of the masses, I’d cut diameters different in 46 series vs 3-21 or 4-18 lights. Triple and quads use a positive contact ring at about the same place 46 series cells are negative under the insulator. Wrapped batteries also cover this are.
This all said I’m a big fan of the 46 series currently only owning two, much greater runtime. A bigger fan of single cell light, no cell matching needed. For those only needed a few cells Skylumens has them. I’d love to see Simon place his talent into building great 46 series lights. I also understand that a true supply of thousands on hand needs to be there for him to have a chance of a profitable run.
The 757G versions work well when soldered on a 3535 footprint, slightly offset. I put together a light with 8x at 0.6A each, soldered offset on a modified 3535 mcpcb and the beam was good with frosted optics.
Supposedly they work up to 1A with good cooling.
The diodes are off to the side, so they don’t work as well when centered.
1 Thank
Feel free to start a separate thread for this discussion, but I’m very interested in the details of your build, maybe some photos? What was the 8x emitter layout, and what optic did you use?
The 1A current probably won’t be achieved as a flashlight does not have ideal thermal properties, but the tests do show that significant overdriving is possible. I’m particularly excited for the SunLike emitter rated for 3V 1A, which, if things scale linearly, should comfortably handle 2-3A of current. [EDIT: misread units, replace “A” with “W”. It should still have better current handling than Optisolis though.]
But again, high CRI emitters tend to lose some CRI when overdriven. Ideally I’d like to see these emitters in big (like 3x3) arrays with heavily textured reflectors or as mules–the angular tint shift does not seem particularly bad from the tests.
I do hope the centering issue can be resolved if Simon makes custom PCBs for these 3030’s! But frosted or heavily textured optics/reflectors are definitely the way to go given the non-uniform emission surface.
Addendum: in addition to the (supposedly) better power handling, I like the SunLike line a bit more than the Optisolis because it uses the violet die for the full range of CCTs, while Optisolis seems to only use it for 5000K. Also the SunLike has a symmetric footprint that might fit Simon’s PCBs for the 276A UV emitters, or B35AM. The spectrum also looks marginally less spiky.
Don’t feel like making it a build thread right now so here’s the short version.
The light is an Emisar DT8 with the Carclo “wide spot frosted” optic. It is using the 5A linear driver, with the FET disabled in firmware.
The emitters used the DTP pad of the original mcpcb as negative, so I had to insulate the shelf of the light. I did this with an extremely thin layer of nail polish so it’s far from an ideal thermal situation, though there’s thermal paste.
Used nylon screws for the mcpcb as steel screws would short negative (pcb core) to the battery.
I built two of them, one in 5000k with I kept, and one in 4000k which I sold to a friend, and couldn’t build any more because I didn’t have more emitters.
With a custom MCPCB like those for B35AM the whole headache could’ve been avoided, but it’s also quite a lot of money for what’s basically a one-off light.
1 Thank
@Simon, can you consider making a 18650 tube + tail cap or 20350 short tube for M21F, M21H, S21F etc? You were able to create a tactical button cap for them since they have the same thread. I’m currently using an 18650 in S21F and the weight balance is better. A 18650 tube could make it a lot more portable.
Optisolis are very bright in triples, which I drive at 0.7A each. Agreed that the optic and/or lens need to be frosted in order to get a good beam!
As singles, they are still plenty bright.
And as QReciprocity42 mentioned above, it’s nice to have a variety of flashlights with different LEDs that excel in their own ways. My triple 6500K Optisolis is one of my favorite flashlights, despite my preference for lower CCT LEDs.
3 Thanks
These SMD LEDs should be specifically designed for LED lamp.Its packaging structure should not work well with reflectors, My guess is that the light spot will be uneven.
I think it only applies to mule
3 Thanks
i think 3030 MCPCB is more suitable
Yes, but regular 3030 does not have the right offset to focus well in a TIR.
Remember that the die is off to the side, so the pads also need to be moved.
I need to think about this, after all, 21700 is the mainstream trend now.
6 Thanks
TIR on it? What is the angle of the light? bead or frosted?
With it comes to high CRI applications I consider reflectors in general to be inferior in mixing the beam to produce the intended color. As NeutralFan above pointed out, these emitters could work wonders in a TIR setup, like the S21D! The beaded TIRs should eliminate the unevenness completely.
For evenness I would guess that the SunLike emitters would be better than Optisolis: despite both having 2 dies, the SunLike dies are centered in the package, while the Optisolis is not.
This is amazing! What triple PCB are you using? And as Simon asks, what TIR are you using?
The problem with Sunlike is the significantly lower max current IIRC. The optisolis beam works very well if the pads are offset the right amount.