LH351D 90CRI Spectrum Tests

The only LH351D ever measured to be not too green is the 2700K version that contactr measured. All the other versions ever measured has a very positive DUV. Usually 0.004 and above, which is pretty high on the green side. Hopefully contactr will identify potentially non green versions with this test. Thanks to contactr for doing this!


This one is bad. 0.0040+ Duv with all optics basically. CCT is a little over 4K with reflector and a little under 4K with TIR.

I’m not even going to waste my time making a chart for it i’m going to just move on to another bin/LED.

I’ll be suprised if there’s a good bin of the 4000k at all. Everything I’ve heard is that the 4000k LH351D is very green and should be avoided. The LED4POWER guy said he stopped selling it for that reason. Yet for some reason, the “dogfart” 5000k is pretty good and no one sells it except arrow (dead to us) and Digikey. I wish LED4POWER and MTN would get with the times and stock it.

Next up: SPHWHTL3DA0GF4UPP6 (3500K)

This is the higher flux bin of 3500K. Seems pretty nice with the TIRs. Even with low power it’s only a little above BBL. Since this is a bit lower CCT even the reflector having higher Duv doesn’t look too green, it starts looking more yellow than green. Will be interesting to compare this one to the lower flux bin 3500K I also have.

That is a nice one! TA suggested to use these above BBL leds with a FET-only driver, but this one is already nice with a 6x7135+FET driver (so without the single 7135 channel), at 2.1A the efficiency is a bit better than FET only. Small detail is that such driver does not exist, someone needs to make one with the correct mode spacing for the combination of 6x7135 and FET.

I put 4000k LH351D to my fw3a. No green at all. Very pleasant tint.

This one looks interesting.


This one is Duv 0.0025 to 0.0007 with Frosted Carclo and 3347K to 3378K CCT. So slightly higher Duv, slightly lower CCT and 1 less flux bin. Overall, it’s fine but the higher flux bin 3500K is better so no real reason to consider this one except for a slightly cheaper price or 100K less CCT.

SPHWHTL3DA0GF4RTS6 (5000K) - purchased August 2018 from Arrow, NOT Digi-Key

These are the ones I have sliced like ~20 of. I thought I had purchased some from Digi-Key more recently but it appears not. Maybe Bob can tell us if the Digi-Key ones are better or worse than this otherwise I will eventually get the Digi-Key one on a future order.

I enjoy the LH351d, its slightly pink or peach, compared to cooler lights
not pink compared to a 9080 219b, but its nice, without green tint

LH351d turns on at 21 lumens in a Tool aaa 2.0 w pebble tir… dont know how that differs from stock

also, as moderator posted above, his triple LH351 gets to 1000+ lumens (driver may be boosted)

my triple sw45 with a stock 2019 RRT-01 driver, that claims 950 lumens w XP-L, gets to 410 lumens, behind a carclo narrow clear with glass lens on top

all that to say moderator is getting similar lumens from the LH351d as the stock XP-L in the 2019 RRT-01


as far as the 9080 219b 4500k, here is the difference between the sw45 in the middle, and sw45k

that sw45 looks slightly pink to my eyes, compared to the LH351d that looks very slightly golden by comparison, this photo gives some idea:

So this is a LH351D 5000K that you dedomed and swapped in yourself? If so, how did you dedome it?


This one is pretty decent but the 2700K and 3500K both have a slightly better tint (regarding proximity to BBL). This is one better flux bin than the 2700K and the same flux bin as the 3500K.

I got the light as a gift from contactcr

I sliced that one. I used 3 washers, the final one being a 0.9mm thick 19mm outside diameter very smooth stainless washer I found at Lowes hardware. With Feather Double Edge razor blade. The first two cuts with the thicker washers then it usually takes 1-3 slices with the final washer depending on how thin and even you want it.

I always reflow them on a 20mm star with no solder on the LED pads so it sits flat. If I want to do multiple at once I also have a 33mm 4XP noctigon with no solder on LED pads too.

So after all this testing I think the stand out ones are 2700K, 3500K if you want to keep the dome on.

For slicing I think any of them with Duv of 0.0035 or less would work nicely depending on where you want the tint to end up.

I think I will make a larger order of 3500K and also get the 5000K dog fart from Digi-Key and hope for the best. Still searching for a decent 4K. I have one lead on an unknown bin from China that I may be able to get a sample of.


Just some prelim testing w/ Carclo Frosted:

CCT = 4030K (Duv 0.0036)
CCT = 4025K (Duv 0.0033)
CCT = 4041K (Duv 0.0029)
CCT = 4054K (Duv 0.0020)
CCT = 4067K (Duv 0.0011)

shaved LH351D.

Modded Tool AAA:
CCT = 4490K (Duv –0.0031)

Color Rendering Index (Ra) = 94.2 [ R9 = 69.5 ]

I like this TIR LH351d light,

it has earned a spot on my mouse pad,
The tint is good during the day

and being twice as bright as my sw45k, cant be bad

respect the 351

of note
all the other lights in my edc rotation are 219b
this 351 is the only outsider to make the cut
I particularly like the pebbled tir beam, very even, useful at close range

efficiency test
with a depleted Eneloop AAA at 1.2v in the Tool that contactcr built w LH351s I got 55 lumens on max.

The same battery in my copper tool, powering an sw45k, gave me 25 lumens…

so yes, the LH351d Is indeed twice as bright as the sw45k… wow…

Looks like MTN has restocked 4000K +90CRI LH351D’s (bare LED). Anyone know if these are the same as the Digikey ones? P6 flux bin, doesn’t say what CCT bin, I’m thinking it might be the same TP quarter bin.

dont know the answer to your questions, but here are some visuals, for reference
the light in the middle thanks to contactcr

max output reference:
the Cu eTool makes 110 lumens with a 4000k 219c or with a 6000 xp-g2, and it is supposed to make 80 lumens w a 219b 4000k

I was checking to see if there was an obvious difference in “efficiency” as expressed as max lumens on Hi. Albeit the hosts are not identical, they all came suprisingly close to the same outputs, on my light meter, calibrated to an HDS:

73 lm 219b 4.5k Cu Tool w reflector, very pink during the day

72 lm LH351d 5k Aluminum Tool aaa w tir, very neutral during the day, but looks pink compared to optisolis and looks yellow compared to 219b… iow tint is relative

67 lm optisolis 6k JetU w pebble tir, slightly green during the day

Where did you find a tir that fits the tool? I have looked from time to time…