The Fabled Nichia 219B - still relevant?

Maukka’s numbers on the ROT66 with the SW45K R9080s confirmed that they have tint shift with increasing output. They start out at like -.02 div at lower currents (somewhat rosy) and grow to –0.8 div at max output (very rosy).

The one and only benefit I saw of the SW45K R9080 over the 4000k SST20 was a little less ting shift within the beam itself (on ROT66). The 219B was very even throughout, while the SST20 had a little cooler corona and more rosy spill compared to its hotspot.

Filters can just be slapped betwen the optics and glass lens on optics lights.

The ROT66 pushes the 219B very hard, that there is no thermal meltdown it does not even run the FET on 100%

everyone here tells you don’t push it hard the emitter is fine at 2-2.5A

I dont know the answer to most of your questions.
they have generated an interesting conversation.

I presently EDC a single sw45k LED in a flashlight with a reflector and clear lens.

I choose different LEDs depending on the CCT I want. I think the <4000k E21a is a better option than 219b or 219c. My priority is not to have tint above the BBL, at any power level.

I avoid flip chips like the plague, due to tint shift across the beam, in a reflector

imo flip chips should only be used w pebbled TiR, a lee filter wont fix an inconsistent tint across the beam, diffusion film wont either.

I do like having a Lee filter option, but it did not help my Olight S1 w High CRI XP-G3 flip chip. It made the hotspot way too pink when it was strong enough to kill the green in the spill. That light uses a frosted TiR, fail. Even when a flip chip tests below the BBL, as HDS reports for their XP-G3 offering, that does not mean there is no green in the spill. They dont test tint across the beam, only the average tint of the entire output.

I respect Geuzzz opinion that the SST-20 may be useful below 4000, no experience

I agree with LightBringers opinion that the N219b is consistent across the beam at any given brightness, unlike a flip chip, that is NEVER consistent across the beam, at ANY brightness. And I respect his opinion that the LH351D may be worthy of consideration. I also agree that the 219c tint, above the BBL does not work for my spoiled self, as well as my sw45k.

I have another flashoholic friend making similar recommendation of the LH351D, specifically for a 5000k dedome, making it “almost” like a Nichia. The LH351D is brighter, but that is not a priority for my use, which is mainly close range, indoors, and less than 200 lumens. That friend is presently choosing to carry an sw45k single in a reflector with clear lens. No Lee filter, no diffusion, just pure pink goodness.

maukka has shown that tint shift across power levels does not produce tint above the BBL with the sw45k. That is also my experience, no green, no matter how low I go, and I often use less than 20 lumens.

To make a decision on which LED to use, it depends on priority ranking of several criteria.
for example

max lumens does not favor N219b, that is its weakness in a marketplace that emphasizes lumens over CRI, at the expense of Tint above the BBL.

choice of battery size impacts expectations of max lumens, I dont use 18650, nor turbo, on my lights.


my preference for small lights comes with a preference for small lumen levels (low power levels), which is a good match for my choice of sw45k. I see no reason to change that atm, given my current use patterns and preferences.

For the record: is the 219B not also flip chip design then? It has no bond wires on top.

The fact it doesn’t shift above the BBL doesn’t negate the fact it shifts a lot below it.

The dedome LH351D 5000K tracks amazing close to the BBL across a huge range. It does have a lower R9 value though.

Re: flip chip or not. Isn’t the 219B flip chip? The SST-40 has terrible shift across the beam in a reflector and it isn’t flip chip so I don’t think it’s safe to categorize this as good or bad. Maybe you mean it’s better for a reflector cause they are generally better for throw?

The 8th minus green filter gets the 219C’s very close, if not right on.

Plot twist, the LH351D is a flip-chip IIRC. I think some of Nichia’s models are too. Both of those have less tint shift across the beam than anything but the HI emitters from Cree. The latter’s flip chips are truly awful behind anything but the floodiest lens possible though.

thanks for helping me learn more
If I was mistaken to blame tint shift across the beam on flip chip design, I retract that distinction

thank you also for helping me learn more
at the power levels I use my sw45k I dont see a problem with tint shift at different power levels, and I dont see tint above the BBL at any power level… thats my preference, NoGreen tint.

I also see no tint shift across the beam of my sw45k, whose max output is 400 lumens, at any power level I use

thanks for helping me learn that even non flip chips can have tint shift across the beam, in addition to tint shift at different power levels

bottom line, I dont want an LED that has tint shift across the beam

I dont like the tint above the BBL of the 219c, I dont like the tint shift across the beam of the XP-G3, and I have avoided the SST-20 due to consistent reports of green tint at low power levels. I use low power levels.

I avoid green tint, whether due to low power, or also when it appears in one part of the beam but not consistently across the entire beam.

You can always get a more aggressive minus green on a competant emitter like the LH351D/SST-20/E-21A/Optisolis to give it more or less rosieness based on preference to near match the 219B at lower levels, and you get better runtimes and less heat and the option for higher output levels beyond the 219B that last longer with less heat to. It’s great for small EDC hosts whose thermal mass is otherwise insufficient for the luminus power flasha-holics desire. I understand there is a desire for going filter free, but you are greatly limiting
your options and platform flexibility by insisting on going without.

thank you, seems I may have caused you to misunderstand my opinion of Lee filters.

I like Lee filters. I use them on several of my lights.
Lee filters did not work on my XP-G3 Olight, due to tint shift across the beam

I got 50x of sw45k, now time to build!

Ah, I thought you were a hardcore emitter purist—no after addititves.

Urg… too much to quote individually.

I think the problem isn’t necessarily with flip-chips themselves (except Cree), but how the phosphor layer’s laid out. Too much slops off on the sides, and then side-throwing light is yellower and straight-out is bluer.

I’m guessing the 219s and (especially) 351s have phosphor laid on more carefully?

My LuxPro lights all have G3s in ’em, and they have tight hotspots like a C8 if not P30, but a hideous urine-yellow corona that almost doubles the “hotspot” size. Dunno what film I’m specifically using, but on the 4×AAA and 2×AA lights, the beam is one smoooooooooth blur of light, no shift, fried-egginess, nothing. Almost as smooth as a wide-angle TIR. The 2×D light I think has a “grainier” film that still shows something of a hotspot, but not nearly as bad(/small) as the original.

As far as non-flip-chip Crees, the ubiquitous 4C XM-L[2?] in S2+es has a fried-egg beam, but cleans up beautifully behind a dropped-in TIR lens. For something nominally above the BBL, it’s actually quite nice.

And any tint-shift is lesser and more gradual, only really showing up when contrasting reflected vs straight light. The G3s show up horribly even when looking at the bare chip (as a mule).

Unno, that’s all I can remember for now.

I modded the newest generation of Tool AAA with the honeycomb TIR with a shaved LH351D. The honeycomb TIR made it go even lower and only dropped to 4500K vs the Sofirn C01S I did that was 4000K and 0.0 Duv with a similarly shaved emitter.

Modded Tool AAA:

CCT = 4490K (Duv –0.0031)
Color Rendering Index (Ra) = 94.2 [ R9 = 69.5 ]

I forgot what I sent you but I've seen an XPG-3 beam go from absolutely nasty to very nice .A 4AA Aldis grocery store light .

Worst tint shifted nasty yellow corona suddenly became 1000% better .

I think not just crappy tints are improved I've seen my best /favorite nichia lights get nicer beams with diffuser film ..I think the #7 frosted film #2 or #4 are actually better than the original Dc-fix when it comes to beam quality or tint mixing .

Forgot offhand the numbers, but it was mix’n’match pieces, some sticky, some not, one thicker like a playing card.

The biggest piece I was able to stick on the 2×D LuxPro light (~43mm, same reflector/glass as a C8). That was the grainiest one (#1?), which smoothed but didn’t blur the hotspot. Still, infinitely better for close-quarters.

The smaller pieces could only really fit on the 2×AA and 4×AAA lights, 24mm and 32mm, I think. Leftover piece I stuck on my SP10. :laughing:

The nonsticky one I samwiched between reflector and “glass” (plastic, long since replaced with real glass). The sticky one I slapped onto the inside of the gla– plastic. SP10 I can’t open, so sticky on top.

Those were well-frosted, really smoooooothed out the beam into a big blur. Again, infinitely more useful for close-quarters, especially such small lights that ain’t throwers to begin with.

Yeh, just like the LuxPro lights. Amazingly tight hotspots for such small lights, but befouled by the piss-yellow coronas doubling the size of the hotspots with Pure Ugly. Add film, and niiiiiiiice.

Those are probably the less-grainy ones, yeh. Way better smoothing.

I have the sw45k and a shaved LH351D 5000k in two identical lights and to be honest, I can’t pic which one I like best. I lean ever so slightly towards the 351D for the tighter hotspot and higher output. But that doesn’t have anything to do with tint or cri. From the many times I have shined both on the wall for comparison, if you like the slightly warmer tints then go with the 351D. If you like the pure white tint then go with the 219b. Either one is a good choice in my opinion.

“pure white”

I get cringe-y when people use that term. I think of the marketing departments that mislead people into buying illegal headlight kits that sell their lights using cockamemy terms like “pure white”, “xtra white”, and so on.

Yeah everything is relative.

I’ve got a few 219B and C. lights.
The Astrolux AO1 Has a B in one, and a Cree in other.
The Cree ranges further but is “cold”light to my eyes.
Same with my S41’s. One Quad “B”, One Quad Cree.
A Cold colour, ranging further in Cree.
but NO colour in Greenery.
I have a coupla others too.
ALL fixed and no problems (to MY eyes) with colour tints.

The 219B just the nicest beam of light to use Personally.