I love these type of simple experiments! Great picture :-)
I understand that the MT-G2 utilizes a variety of color tint emissions in all those tiny die surfaces to control output. Is this why it’s so clean and buttery smooth? If the XHP70 and 50 utilized different tint bin dies like ToyKeeper prefers, it might look a bit odd at a direct glance but the ensuing output would, potentially, but a much higher purity for better CRI and smoothness.
How cool (neat, premium, superb) would that be? :bigsmile:
Yeah, I’d say that’s the case. With the mt-g2 die surface, at any one time there are far more emission angles taking place than with a typical flat die. Looks to me like the light has a big chance of bouncing around between those little phosphor turrets and being blended towards a smoother tint.
Aren’t there MKR emitters that already do that? Mixing different dies for better more consistent output? Wouldn’t surprise me if Cree are already doing that with the XHP70 as well to achieve those tighter 3step Macadam bins. Can’t see any visual difference between the dies in my particular batch however.
It wouldn’t help much with the colour over emission angle issue.
I’m wondering about a de-dome on the 70, has anyone done that? Not sure what I’m thinking there, it’s not likely to throw any further and the emission would just be 4X messed up, right?
C12476_MIRELLA-50-S was the only one that had a defined spot. The others were more like a flooder.
Mouser has the 90 CRI 5700K XHP50 in stock now in case anyone is interested.
The XHP’s come in two versions, the single tint like 1A, 3D & the grouped version called E3, E2, E4, E5 in for example the E3 it is one die of each 3B, 3C, 3A, 3D.
The pure single hue versions can be found in higher flux bins, if you want it with an easywhite bin like E3 group you often lose a couple of flux bins.
I think, perhaps, someone has misinterpreted the data sheet.
The group you’re talking about encompasses the chromaticity of those 4 tint bins, it does not include one of each. You are taking a gamble when buying a group binned emitter as to just where in that wide area your particular emitter will be. The only way to ensure you actually get the light color you want is to buy, specifically, the 1D or 3D or whatever tint bin if it’s available. The XHP series, to the best of my knowledge, is not actually rated as 1D or such, those bin numbers have been added in as they’re being sold. You can get an N2 or N4, for output binning, but unless each emitter is being tested by the seller and grouped into the A,B,C or D category they really aren’t being sold that way.
Just like how there are sellers touting XP-L V6 1A… there is no V6 1A. They are selling V5 1A that have the highest outputs, presumably, as Cree has not released a 1A designation in V6 binning so far as can be determined as of relatively recently.
Certainly, if someone is measuring the color of each emitter before selling them, then they could group them into those classes, but you don’t order from Cree that way unless I’ve missed something somewhere. (wouldn’t be the first time) I am assuming this is due to it being an early release and that these specific tint bins will start becoming available as the series gets some age on it.
Please feel free to set me straight, somebody, if I’ve got this wrong.
Well we have definitely seen Cree offer new output bins that weren’t in the datasheet over the production life of an emitter. Nothing unusual about that.
The cree datasheet gives me a headache, but I think there’s a real distinction between the classic chromaticity bins we’re used to (3A 3B 3C etc, both individually sold or in broader groups such as E3) and the Easywhite stuff.
Individual bins seem to only become available after a while, initially the broader groups are what’s on offer. With the XHP70 though, just as with the Mt-g2 there’s also the Easywhite bin that’s available.
*or 3 step apparently when it comes to the XHP70s 
This is obviously a method only possible with multi die emitters. So specifically with the ones I ordered from mouser I would indeed expect my emitters to be a certain combination of 3A 3B 3C and 3D dies to achieve a combined output inside that 3 step macadam ellipse.
I red the entire datasheet, multiple times, as well as the announcement white paper and all that stuff….How did I miss that? Geesh, getting old sucks!
Thanks for the clarification Linus. Sorry I’m so dense Cajampa.
Anyone else ever use the “baking cookies” analogy to explain emitter binning?
It’s all about fine tuning those sprinkle dispensers 
Are they using Pop-Tart sprinkle dispensers for applying the phosphor? Looks like it….
Makes me wonder if we couldn’t track down some “red phosphor” powder and sprinkle a bit extra on to counteract the green shift of an average dedomed led………… H)
……I’ll shut up now
No worries Dale 
@Linus, Yes you are right there are 3 different tint sorting systems they use, i was wrong on that part, the grouped & the 2,3 & 5 easywhite step eclipse is very different, & i just finally got how, i hope
Check out these images & i think it makes much more sense.
As far as i understand this easywhite various step MacAdam tint sorting is, they are sorting them after how close they get to the BBL, and not as a traditional ANSI tint at all, i think that the best premium sorted dies that gets really close to the BBL (thats what the coordinates signifies in the pdf page 16 for example) like a 2 step Easywhite sorted tint, is used in applications where many leds are installed and there can be no differences between led’s or drift over time in hue from the other led’s in the installation.
The “E” group version is for ANSI spectrum mixing & probably work best & look best in a single led installation or at least where it is not so color critical.
And of course our standard ANSI-white sorting with some nice favorite tints & many that are not so nice in my eyes anyway.
EDIT
This page explain the MacAdam step sorting principle better than i could also
And i gather that if it is 1 step or 2 step it basically is experienced as pure white, only in different color temperatures without the green bias from the above BBL tints or the pink bias from the below the BBL tints.
That would certainly be the only way to sort single die emitters, but for multi die or array type emitters it’s also possible that they grade each individual die separately and then intentionally combine them in complementary ways to create a final quad emitter with the required tighter Macadam grading.
e.g Combine a “green” 3B/3C die with a “pink” 3A/3D die (as many have done themselves) and you end up with a blended output right on the BB line.
The rather wide 3 step macadam rating may be a symptom of them not being able to generate enough of the pink emitters to properly balance the more common greens. Pushing things on average more into the green and stretching the ellipse as a result. Pure speculation of course, but the A and D bins certainly seem to be rarer that the B/Cs when talking about XMLs.
No idea if they actually go to the trouble of doing that with the XHP70s or if they just randomly assemble emitters and then bin them afterwards as you say.
However I don’t think I’ve seen Easywhite binning used to describe any of Cree’s single die emitters, but I could be wrong on that.
I haven’t seen any single die cree led’s in easywhite either, but i would love a tight 1 or 2 step NW 5000K XP-L
But unfortunately because we haven’t seen any single in easywhite maybe it can’t get that tight, maybe it is just that precise sorting + mixing that gets them that close to the BBL at all.
Buy enough emitters in a 3A/3B/3C/3D grouping and you’re almost guaranteed to eventually find that special emitter that will be bang on the BBL, unfortunately they don’t seem to be common enough to be individually selected or binned. Actually to have any chance of getting one that’s exactly on the BBL, you’d have to sort through only the rarer 3A and 3D emitters since the other two bins already start right above the line. Good luck!
But even then there’s still the Cree rainbow to deal with so it’s pretty much always a compromise.
I just did a 4 x AAA light with an XM-L2. It is great, except for the 20 minuet run time. The Eneloops get hot but not unreasonably so.
So perhaps 3 x 10440 would work well for a 6 V LED. That could fix all those boring 3 x AAA lights!
In case anyone wanted to know. Mouser got thirty 90CRI 5700K XHP70 in stock today. XHP70A-01-0000-0D0UK40E2
What do you mean by taste the rainbow?
Do you think that cop has a CREE XM-L Color LED in that flashlight?
Like this one?
http://intl-outdoor.com/cree-xml-color-led-16mm-mcpcb-p-724.html
That might be cool to have a rainbow light but I wouldn’t want people to get the wrong idea about me.