Nichia 319A D440 Output test by Texas_Ace; Hex die LED gives a nice beam

Ok, like normal, tests done the same as all my others.

This is the new Nichia 319A LED with the new Hex die. I put it through the test and the results were not quite what I was expecting but also not that far off.

Marginal improvements over the 219C in a similar spec. Large die means less throw and more flood if that is what you are going for but it also means a nicer beam pattern, particularly on optic based lights.

I think it might be thermally limited as when compared to the 219C it falls off quicker. It is quite possible they will release a B and C version of this LED in the future with thermal improvements and then it could really shine. The smaller thermal pads on the nichia LED’s really struggle with high power outputs compared to the cree pads from what I have seen.

Overall a nice LED, just not a game changer as some of us hoped.

Click images for larger version:

Compared to a D320 219C:

Great info! Looking forward to building a 319 triple!

1151 Lumen @ 11.5 Watts is pretty good.

Yes,t he Vf is much lower so the overall efficiency is much improved when it comes to lumens per watt.

So… marginally higher lumens (55lm at most) until about 6amps. And a bit lower vF. Eh. It’s good to have options though!

Thanks for the test TA!

I see a great led for a parallel triple on a single li-ion cell, compared to the 219C a slightly smoother beam with more output because of the even lower Vf.

Thanks for the test. This falls a little short of what I was hoping it would do, but all the applications I have in mind for the ones I ordered were more throwy than I wanted with a 219C, so I’m happy to have a solution to that problem. These should make great triples. I wonder if an R9050 option is coming.

Like I said earlier, this is about as bright as the D320 219C.
A quality upgrade for 219series in the near future, but not yet. Nichia cram more emitting area with approx. the same thermal pad footprint. I guess 319C D600 is to be expected in 2019, who knows?

Tex, greatly appreciate the test!

- Clemence

My guess is that the thermal pad is the limiting factor even on the 219C, so it makes sense that the 319A doesn’t make drastic increases in output. I wish they would at least adopt the cree footprint with the slightly but noticeably larger pads.

I really wish that the 5050 footprint would be updated. It has so much more thermal shedding ability I bet that existing dies would get a good boost in performance from that alone.

Contrary to what we would expect:

- 119A/B/C and 144A is Nichia answer for larger thermal shedding performance…… :expressionless:

  • What is Nichia recommendation for flashlight applications? The answer is this: SERIOUSLY??

:question: :person_facepalming:

Nice work TA. Invaluable info. :+1:

I think what you meant to link was this

It’s mostly 5mm strawhat LEDs, but this looks pretty acceptable.

Yes, E17A and E21A is MORE than acceptable. It’s another “219B/C HI” with wide range of Ra9080 [2700K-5700K]
No I meant, I literally asked one of the Nichia guy and also read at the hardcopy catalogue. And the most “flashaholic acceptable” answer were some DIP and NCSL types.
Different standards…

Like an XP die on an XM substrate?

I was checking up on the E21A, it looked interesting but then I noticed there’s no thermal pad. Kinda puts a damper on things, no?

Just like those 144AM/AR :frowning:

Yes, except for AAA lights there is basically no situation in a flashlight where a 5050 footprint would not fit and it offers a much much much larger thermal pad which can only be a good thing.

Plus 3535 LED’s are generally rated for a max of 2A where 5050 generally get rated for 3A. Obviously showing an improvement in thermal path to some degree.

Exactly how much of an effect it would have, I am not sure but I have seen a consistent limit of between ~35-40W for the Cree 3535 LED’s and around 20-25W for the nichia 3535 LED’s.

Having seen this on many LED’s, the results would seem to lead to something besides just limits within the die itself and more in the thermal path.

Much like the first and best place to get more traction in a car is wider tires, I feel that starting with a larger thermal pad is the first thing that should be done with LED’s.

But then I am the guy that will cram 325’s on an MR2 and then run drag radials on top of that to multiply traction even more. (for those not car guys, this is roughly the equivalent of a 8x XHP50 SRK making 17,000 lumens in the flashlight world, not completely unheard of but WAY overkill).

Did you see what Vinh just did with an Acebeam X65 and 28 XHP70s? No, that wasn’t a typo. Twenty eight XHP70s.

I prefer dependable “practical” modded lights. Don’t you think it’s better for him to just use COB? More compact optics too?

I don’t pay attention to him myself.

I will admit that I am bummed he was the first to make a true 100k lumen flashlight. I have actually been working on a setup that should do this for this years OL scratch build contest.

Seeing as 100k lumens would be around ~1200-1400W and at most would run for ~10 minutes off 2x massive lipo packs (and yes, my planned build should be able to run the full ~10 minutes at roughly full power without overheating), anymore then this would be a bit impractical for more then a 30 second burst but mostly due to shear size it would take for the optics.

Any details on the Vinh light? Curious how he is cooling it. It will be producing about the same amount of heat as a space heater, heat gun or microwave.