testing two of KD's new Nichia 219b leds (4500K'92CRI' and 5500K)

[malkoff_junkie]

so which one out of the 219A and 219B will give just a pure white tint?.

219B?.
is there a name for this actual LED for pure white nichia or is 219B ok?.

thanks.

[djozz]

@zebrastripes

so which one out of the 219A and 219B will give just a pure white tint?. 219B?. is there a name for this actual LED for pure white nichia or is 219B ok?. thanks.

Difficult question, I can not answer that, there's no such thing as pure white, every tint appears to have a slight colour relative to every other tint. There is however a favorite tint, and that one differs for each person.

[malkoff_junkie]

are there any lights out there carrying this like white 219B LED in 18650?.

thanks.

[djozz]

@zebrastripes

are there any lights out there carrying this like white 219B LED in 18650?. thanks.

No, not yet at least, it usually takes months or more than a year for new leds to reach stock lights, and only if they are significantly better than what what they are used to.

[Lithium]

Thanks for the tests

@djozz : where do you found the Osram 96 CRI ? Is this this one : http://www.led-tech.de/de/High-Power-LEDs-Osram/Osram-Oslon-SSL/OSRAM-Oslon-SSL-80-neutral-weiss-CRI-95-auf-Star-LT-1956_206_207.html

if yes, is it well suited to be put in small flashlights (considering 800mA max) ? It is a bit smaller than XP-G, but not that much, how could I have missed this led :quest:

Edit : pulled the trigger on some 219B at Kaidomain, wanted to see by myself 0:)

[djozz]

@Lithium

Thanks for the tests @djozz : where do you found the Osram 96 CRI ? Is this this one : OSRAM Oslon SSL 80 neutral-weiss CRI 95 auf Star - Oslon SSL 80 LCW CQ7P.CC-KQKS if yes, is it well suited to be put in small flashlights (considering 800mA max) ? It is a bit smaller than XP-G, but not that much, how could I have missed this led :quest: Edit : pulled the trigger on some 219B at Kaidomain, wanted to see by myself 0:)

That is the led I tested, although I bought it from rs-online and the test was done with the led reflowed on a 16mm Sinkpad board. I have used it in a Ultrafire M5 (small AAA light) and in a sk68-clone (AA-zoomie), I described it in that thread. Of course the amount of output was not all that special (but 96CRI, and it shows!), but because of the narrow emission angle it behaves very nicely in the various optics, in the sk68-clone the light loss is even less than the XR-E.

[Lithium]

Thanks, I’m going to order a few, to try along with the 219B
But I have no reflowing skills, guess I’ll have to stuck with 16mm boards.

[djozz]

@Lithium

Thanks, I’m going to order a few, to try along with the 219B But I have no reflowing skills, guess I’ll have to stuck with 16mm boards.

Under 1A there is not much to be gained with a copper board anyway :-)

[Helios1 (Helios-)]

Reflowing is easier then it may seem. Just try on a cheap led.

[malkoff_junkie]

BUMP.
thanks.

@zebrastripes

so which one out of the 219A and 219B will give just a pure white tint?.

219B?.
is there a name for this actual LED for pure white nichia or is 219B ok?.

thanks.

[Lithium]

@Helios-

Reflowing is easier then it may seem. Just try on a cheap led.

I already tried, guess I’m going to need a better soldering iron (mine is a cheap crap, I already planned to buy a new one but can’t found something that screams “quality” locally, haven’t really looked on the Internet for now I admit), because I failed miserably.
I wanted to try in the oven too, but my wife totally disagree on this one :bigsmile:

[djozz]

@zebrastripes

BUMP. thanks.

@zebrastripes

so which one out of the 219A and 219B will give just a pure white tint?. 219B?. is there a name for this actual LED for pure white nichia or is 219B ok?. thanks.

I see, well, it is difficult with Nichia, the 219B comes in various tints and CRI. But if Cree's coding system is complicated, Nichia is even worse. I could not find any documention that directly couples the exact specifications of a particular led to a order code. It does not help that Kaidomain is sloppy with their descriptions of codes either.

I love Osram and Philips for their clear and to-the-point documentation.

[malkoff_junkie]

thanks.

[RedForest_UK]

I think the 5000k Luxeon ‘daylight’ tint that Zebralight currently offer in a few models has been stated to be close to a ‘pure white’ tint. Remember also that there is no such thing as ‘pure white’, our eyes adjust massively to varying colour temperatures in the environment and what may look rosy tinted against a white wall with cloudy daylight coming through the window may look green tinted on the same wall when the room is light by incandescent.

The proportion of R/G sensitive cones in the retina can vary between individuals too, I posted this in another thread a while back:

@RedForest UK

In a 1999 study by Roorda and Williams; ‘The arrangement of three cone classes in the living human eye’, they applied relatively new ‘Adaptive Optics’ previously only used to reduce blur in telescopes to give an unprecedentedly clear ‘map’ view of the ‘mosaic’ of photoreceptors in the fovea after selective bleaching of Long, Med and Short wavelength sensitive cones .

They found a massive difference between two subjects, both with apparently normal colour vision. Subject one ‘JW’ had 75.8% Red cones, 20% Green and 4.2% Blue, while subject two ‘AN’ had 50.6% Red cones, 44.2% Green and 5.2% Blue.

The psuedocolour images in this PDF write-up of the study at the top of the third page are particularly striking: http://vision.berkeley.edu/roordalab/Pubs/ROORDA.PDF

In my lectures it was implied that this high level of variance has since been shown to be common within any given population, although of course ‘AN’ could have been abnormal.

They concluded:

‘The large individual differences in numbers and arrangement of
cone classes that we have observed indicate that evolution has not
converged on an optimum proportion of M and L cones for the
human eye. Is this because red±green colour vision is a relatively
new feature of vision in old-world primates, or do the statistics
of natural scenes, optical blurring and clever post-receptor processing
make M- and L-cone topography unimportant for visual
performance?’

So there is likely a lot of individual variance as well, whilst to some 4000k might look like ‘pure white’, to others 6000k might look like ‘pure white’ in similar conditions. You have to find your own general preference and look for that, mine is generally around 5000k i.e 3D/4A.

Btw, as that old link is now dead, here’s one that works if you want to read the paper: http://aria.cvs.rochester.edu/papers/Roorda99nature.pdf

[djozz]

@Lithium

@Helios-

Reflowing is easier then it may seem. Just try on a cheap led.

I already tried, guess I’m going to need a better soldering iron (mine is a cheap crap, I already planned to buy a new one but can’t found something that screams “quality” locally, haven’t really looked on the Internet for now I admit), because I failed miserably. I wanted to try in the oven too, but my wife totally disagree on this one :bigsmile:

If you plan to do many reflows in the future (of course you do!) it is worth the trouble to build a set-up for it. If you search BLF for 'reflow' there's many nice ways to do it.

I really like my way, a rebuilt 12V solder iron with a aluminium block mounted on it, it took some effort but I am still really glad I made it. I described it here some time ago (please do not watch the video , I am very handy with reflows nowadays). It is compact and easy to work with, and after some experimenting with the best voltage, it can be set to a constant temperature not too far above the solder melting point, to minimize led damage.

[Lithium]

Thank you, I’ll try to build something like that

[syracuse]

I’m one of the lucky man who had got 5 pcs of 219B from IS a few weeks ago. Then I made a triple 219B on Donn’s solid copper heatsink together with DrJones’s lucidrv driver. The result was impressive… (I do not have 219B from KD to compare)

The triple 219B is considerably brighter than triple 219A (same set up). The B render BLUE and WHITE better than the A. On the other hand, The A render RED and YELLOW better than the B.

Sorry for my crap photo. Both Nichia make the triple XPG2 R5 3C on the left looks yellow. Triple 219B is in the middle. Right hand side is our familiar pinky 219A which is still my favourite.

[djozz]

Thanks a lot Syracuse, that is very useful information. Actually your observations on the 219b high cri from IS match quite well with what I found from the 219b high cri from Kaidomain. It let me suspect that KD was not selling the high cri 219b. So now every scenario is open:

1)KD is right, and so is IS, and I did not recognise the 92cri because I like warm tints too much.

2)KD is wrong (not 92cri) and IS is right (different tint than 219a but still 92cri), but also

3)both are wrong, and IS suspected that and withdrew their version of 219b.

4) IS is wrong and KD is right? That would be the world upside down , and nothing supports that

("exactly, my dear Watson" )

[syracuse]

At first, I also doubt about the 92cri. But this 219b renders colour much better than the XP-G2 R5 3C. 0:)

I do believe in IS, and I think they use this emitter in their FOURSEVENS Preon P2 High CRI Edition Neutral White.

[comfychair]

I.S. just ran out, they got a limited quantity for the special run of the P2 and had some left over, that's all. I'm sure when they're able to get more they'll be listed again.