Nichia 719A 5000K R9050 tested

Flashlight Modding and DIY Parts
1

I received a request from Simon from Convoyflashlights to test the new Nichia 719A. I have retired from testing stuff for a while, but I found this one interesting enough to do one more (no trend probably). I hope to get familiar with editing posts on the new website too, forgive me for not being fluent with the new editor.

The 719A is an interesting new led because it is a 6V domeless led without any visible signs of being multi-die, it seems one single die, I wonder what is the technology that Nichia uses here? (Edit: thefreeman has some more info in post#3, the trick is a ā€œstacked dieā€)
Here it (the 5000K R9050 version) is at very low current to show the emitting surface:

Here is the page of the 719A on the Nichia site:
And a link to the datasheet:

The page tells you that it is a ā€œsuper high luminanceā€ led, and according to the datasheet, the max current is 1500 mA, pulse current 1800 mA (power-equivalent to 3000/3600 mA if it was a 3V led).
Here it is among some other domeless leds that I had around. The die size is closest to the XP-L Hi. I maybe should have added a sliced (dome removed) Samsung LH351D in the picture, the 719A die is smaller than that.

The leds are:
Luminus SFT-70 Luminus SFT-40 Yinding 5050
Cree XHP35 Hi Nichia 719A Osram CSLPM1
Cree XP-L Hi Osram CSLNM1

Tint measurements (pics see end of post)
I received 2x 5000K R9050 (one for output test destruction), 1x 3000K R9050, 1x 2700K R9050 from Simon, first I did some spectrum testing, for that I placed a smooth reflector on top of the leds and measured in the hotspot produced by the reflector. Btw, I have no phto of that, but the beam it produces is about perfect, no rainbow issues whatsoever.

Because I do not know yet how to resize spectrum pictures and place them next to each other, the (many) spectrum images are at the end of the post and edit them to size later. My general impression is that the claimed CRI is about correct up to 2A, but at 3A it goes significantly down (the output data shows indeed that the led is getting less happy at 3A). The 5000K version is a bit above the BBL (mind that hotspot measuring from a reflector usually increases the duv a lot from measuring the bare led), but the 3000K and 2700K stay very nicely near the BBL.

Output measurements
The voltage/output test was done like all my emitter tests except the really ancient ones, for links to my other led tests, see the djozz-test link when you click my avatar. I described the method in detail in my XP-L test, with two minor upgrades since that should not matter significantly for the results: I used my Integrating sphere no. II instead of no. I, and for the current I used a clamp meter, which appears to measure 0.1A lower than the power supply current-reading that I used for my earliest led tests.

Some further remarks about my led tests: 1) just one led was tested which I hope is representative for the whole batch, reflowed on a DTP copper board (XP-KDlight 20mm) that was clamped to a chunky cooled copper pillar with heatsink paste between, 2) I used my large version II integrating sphere with high quality luxmeter, 3) the output numbers and voltages were measured with the led close to ā€˜steady stateā€™ for each current, so warmed up and settled, you should be able to get these numbers in a well heatsinked flashlight. Mind that these are output numbers of the bare led, in a flashlight there will be losses from light obstructions, lens and optic, 4) output is in ā€˜djozz-lumenā€™ defined as 1/550 of the output of my Sunwayman D40A on high setting, which I hope is close to the real lumen, but at least is consistent over all my emitter tests done in integrating spheres. A few years ago I checked the djozz-lumen against lights measured by maukka and the djozz-lumen seems 7% to 11% high (depending on spectrum) compared to maukkaā€™s calibration.

Here are the results for the output test of the 5000K R9050 version of the 719A, sorry, not a nice computer graph because my ancient Mac does not communicate with the world very well anymore, I may update this post sometime when I got that working again

The sharp decline after 3.5A seems to me a sudden failure of the led, including severe damage of the phosfor. Going back to 2A never restored the output (I got about half the output back).

The maximum output seems on par with the XP-L hi of a couple of years back, Texas_Ace found 1620 lumen at the time, if you take into account that this led is R9050, 1280 lumen max is about consistent with that.

Conclusion
For Nichia to make a domeless led with the approximate luminance of a Cree XP-L Hi is new and special, but the luminance may still be a bit underwhelming when reading their jubilant claim about the technology.
For me as a flashlight modder a 6V led is more complicated to use than a 3V led, but there are some very nice boost drivers nowadays if you want to stick to single cell (like me).

-------ā€-------

Spectrometer readings

719A R9050 2700K, at 1A :

719A R9050 3000K, at 1A :

719A R9050 5000K, at 0.3A :

719A R9050 5000K, at 1A :

719A R9050 5000K, at 2A :

719A R9050 5000K, at 3A :

32 Thanks
2

I hope I will manage to get more than one image next to each other, and that I can set width percentages for images. Iā€™m learningā€¦

4 Thanks
3

Itā€™s a stacked die construction apparently.

https://led-ld.nichia.co.jp/en/product/lighting_shl.html

This is the worldā€™s first LED that achieves super high luminance through an innovative structure with multiple light-emitting layers.

Thanks for testing it.

3 Thanks
4

When the images are uploaded to BLF you can set the % size. (Hover on the preview and the resizing commands show up).

For resizing 3rd party hosted images i think you need to add HTML code.

Edit : That seems to work :

<img src="https://i.imgur.com/ApF5lBO.jpeg" height="200"  /> <img src="https://i.imgur.com/onh7gez.jpeg" height="200"  />

It previews correctly when setting only the width, but not when only the height, it still works though, setting only one dimension keeps the ratio.

And to add a link to full size :

<a href="https://i.imgur.com/ApF5lBO.jpeg"><img src="https://i.imgur.com/ApF5lBO.jpeg" height="200"  ></a>

1 Thank
5

Thanks for the info, thefreeman :+1: I will finish the post first, and then study your tips and edit it to make the post more readable.

6

image
image1292Ɨ994 57.3 KB

Itā€™s with Maukka lumens (*0.93) because thatā€™s what I had in my spreadsheet.

12 Thanks
7

Thank you djozz. I wonder what Nichia planed to achieve with 719A, since performance looks close to 519A dedomed, but with lower r9 and a more complex design.

8

Your results mirror my experience with 4000K 719A this week. These are informal uncorrected measurements taken from the PSU display, with a beaded 5050 optic I use for quick testing. At 4A, the LED started smoking and the phosphor was permanently damaged and increased Duv at lower levels. Around 2.25A it starts to fall below the R9050 spec, so up to maybe 3A how hard you run it is more a matter of how much loss of colour quality youā€™re willing to put up with for diminishing output returns. I think 2.5A is a sensible max current.

In an actual light (the first 719A light?), output is maybe a little underwhelming. H1 here is about 2.8A in thefreemanā€™s driver testing, and H2 is 1.6A (datasheet spec is 1.5A max). I can get results nearly as nice as this from a carefully chosen dedomed XHP50.3, but unlike Cree I trust that most if not all LEDs on this reel will have similarly neutral tint. And they donā€™t cost $6 each.

It seems like the ā€œSC64c HI Nā€ or whatever the 719A version ends up being called will probably have pretty similar output to a 519A SC64c LE.

7 Thanks
9

Did you get any candela measurements with your ZL w/ 719A?

Iā€™m curious what the cd:lm ratio is compared to XHP35 HI or 519A dedomed. I guess it would only make sense to compare if the H600 OP reflector is the same as the SC64 one.

10

Itā€™s pretty throwy. I wouldnā€™t consider measurements from this light meaningful since it has a pad adapter and I also temporarily reinstalled the frosted lens for testing. Iā€™ll have a more direct comparison with XHP35 HI later this week.

1 Thank
11

Very interesting results, thank you djozz for the testing. The performance of this LED seems disappointing, around that of a dedomed 519a, while being a harder voltage to drive. Hopefully it will improve with time, but this seems not to have been the XHP35HI (v1) replacement I had been hoping for, with the huge CCT-current shift and lower output.

Any chance of a review of the 219c-v2 domeless LED? It might be a nice compliment to the XP-P in the small but growing category of warm/neutral throwy LEDs.

1 Thank
12

Thank you for the testing Djozz.

I agree, the test results are disappointing. For our uses 519a dedomed just seems better.

13

I still think a factory domeless LED, even if equal performance to 519A dedomed, is useful for those among us who prefer neutral (non-pink) tint and also prefer a more neutral CCT (5000K).
You can somewhat get this by slicing a 5700K 519A but you wonā€™t get the full bump in intensity (candela) unless you fully dedome.

Iā€™m still looking forward to the 719A and also the 219C-V2 (domeless).

1 Thank
14

Ok, I played around with that, and I managed to get the pictures next to each other, and at a certain size (width=ā€œ100ā€, so they appear in one row even on my phone).That helped a lot already, but this forum website seems not to accept dynamic sizing to a percentage of available width, so that the available width is filled out for all screen sizes. I did use the correct html code for that but the % symbol (that should trigger dynamic percentage sizing) is ignored by the forum software.

1 Thank
15

Plus a lot of flashlight manufacturers are simply not going to sell lights with dedomed 519A, especially using the messy dome popping method that makes it practical for Hank, Convoy, FF, etc.

16

So the 719A has lower efficacy (lumens/Watt) than 519A despite having lower CRI and less red in the spectrum! Thatā€™s a huge step backwards.

For anything less than 4500K a dedomed 519A would just be better than a 719A. For higher CCTs Iā€™d much rather take a sliced 519A over a 719A.

2 Thanks
17

yupā€¦

giphy

4 Thanks
18

Thanks for testing this.

19

At 11W itā€™s about -5~-6% for the 719A.
Cree mentions -10% for flat vs domed (same CCT).
I donā€™t know how much 500K and R9080 is supposed to cost but at the same CCT dedomed 519A will likely have lower efficacy.

4 Thanks
20

Itā€™s great to see you posting again, and testing, even if itā€™s just temporary. :heart:

Learning is good too. When the site migration happened, I learned lots of new things I wasnā€™t planning to learn. :sweat_smile:

* I didnā€™t like the look or feel, so I changed it ā€¦ a lot. Now itā€™s a theme users can choose in the chooser widget at the bottom of the sidebar.

5 Thanks