Can you tell them to take macro shot of just one emitter from that reel? Difference can be easily seen…
Good old G2 S4 2B
New production type xpg2s42b
I’ll take 50 or more…
That is 250 pcs so far… Hope Mark is reading this 
Hi luminarium,
All of the XPG2s I had up till now must have either been from the origonal type or I just didn’t notice the change. However I will throw my 2 cents in. This “upgrade” was not only done to XPG2s, but also to XML2s. Or at least “a” change was made. I never had any trouble with a change in tint, however the output from a single XML2 U4 bin dropped VERY significantly about 3 months ago due to a spike in the needed VF. I tested piles of LEDs from multiple sources to try and regain some ampers, but no luck.
I probly would not have even noticed accept I test custom builds like my shorty Maglites before they go out on a regular basis. Out of all the XML2s I have used since then I have found only one that tested out the way nearly all of the ones I got off RMMs first reel last year did.
I just hope when XML3 lands we recover some of that nice low VF as most of my builds use a mosfet driver.
Vestureofblood,
Thanks for reporting. Your post goes in a first post of this thread as a warning that they probably changed production process for XML2’s
You are the man that works a lot with different emitters and especially mentioned XML2s and I worked with XP-G2 S4 2B exclusivity…
When differences are so big that we can actually visually spot it without lux meter and other stuff than we know something bad is happening there…
I hope Cree will make up for this mess with XPG3 or XML3 but if they will remain on new production process? I don’t know…
Did some testing today with one XP-G2 from FastTech (unknown bin) and with these three XP-G2 S2-1A’s from Kaidomain (http://www.kaidomain.com/Product/Details.S023026). Results are pretty disappointing, I was expecting much more, hoping that these leds from Kaidomain are from older batch. Testing was done with Samsung 30Q’s and Samsung 25R’s, both batteries delivered similar results. I wasn’t using any drivers, just direct drive trough 18 AWG wires. XP-G2 from FT is in my Brinyte 158 so I managed measured throw (194 kcd, 30 sec, 3,7m).
1. XP-G2 (cold tint, unknown, from FT) - Bat 4,21V, Max 3,10A, VF 3,90V
2. XP-G2 (1) (Kaidomain) - Bat 4,21V, Max 1,5A, VF 4,08V
3. XP-G2 (2) (Kaidomain) - Bat 4,21V, Max 3,76A, VF 3,82V
4. XP-G2 (3) (Kaidomain) - Bat 4,21V, Max 3,83A, VF3,80V
As you can see, results are all over the place with very high VF. For now I’ll wait for someone to test the new XP-G3 and make my decision on what to do with my Brinyte 158.
I see a rise in 2 cell hosts and buck drivers coming…
luminarium iaculator
There might be a hope in Samsung 3535 high binned emitters, i just got mu hands on some of those binned at 160lm per watt, and initial tests showed their cap could be removed easy
Will check them asa i got home today
Please keep us updated… Fingers crossed…
Edit:
Maybe even LH351B?
http://www.samsung.com/global/business/led/lighting/component/high-power/lh351b
Samsung’s product rendering certainly looks interesting. Nice smooth die and no weak bond wires
FYI, a few weeks ago I checked LG HC35C1 led (=a few years old, the newest gen. is HC35C4, impossible to obtain for me) for hot dedoming and in petrol, in both methods the dome would not come off in one piece and bits of silicon remained over the die. The phosfor was pretty solid though, it allowed scraping off remaining bits of silicon with a toothpick,but not all, tiny bits were left on the die.
Any luck, Mitko?
Getting ready to build a new batch of lights and wish I could get my hands on some of the original XP-G2’s or something similar. If not, I might have to just use the XP-L HI.
Well by default its something like S2, havent tested it in details with the dome on, it is capable of maitaining 4.5amps ( with a single battery and a fet dd driver), dedomes pretty clean though.
unfortunately when dedomed it performance drops dramaticly, its like 40% weaker than those upgraded new S4 emitters( 145kmcds vs 100kcds in a C8 host)…excuse me, but even a dedomed XPE2 performs better
It has a potential, i am pretty sure Samsung will develope those in future but for now those aint suitable for out throwing needs
Better not, try dedomed V6 instead, it outperforms any HI by 25-30%, or even a dedomed U3
Just curious, have you noticed real-world difference in your builds? I understand that V6 should out-perform the current highest binned HI’s (V2 or V3?), but dedoming is going to reduce overall output of the V6, no?
After reading through this thread, I came away with the impression that there’s no real advantage to dedoming a XP-L (as opposed to using a XP-L HI)
Ofc i did m8 especialy considering throw, those builds are mainly used in a real world situations, mainly hunting: many of them, in lotsa different hosts
The mentioned thread is old, just a build that i made today: actualy those were 2 flashlights: both in B158host, one with XPL HI ( v2 3b)one with a dedomed V6 1A, the first gave 170k mcds( at 5.5amps), the second 215k( at 5.6amps)- the reason i made the first one with XPL HI is that fella hunter will swap the pill inside randomly with another pill( ir based) so i presumed HI in that situation would be better, he could simply damage the dedomed emitter occasionaly
Try out V6 1A from MNT or KD , they dedome pretty okay( not like V6 2D or V6 0d from FT) youll see yourself the difference, they “sing” paired with a nice fet dd driver and a 30q/vtc5 battery
Good to know, thanks!
One of the lights I’m getting ready to build is a B158 with the new FET+1 from MTN and a 30Q. Also putting together some UF-1504s with similar components. How’re you liking the B158?
And how do you prefer to dedome - hot or gas/petrol?
Reposted here from my sales thread.
--------------------
OK, here are the results:
"Old" XP-G2
3.87 vF @ 3.9A
"New" XP-G2
3.78v vF @ 3.9A
I tested two samples of each, and they are both within 0.1v of each other. I see that much variation or more from different emitters from the same piece of tape. I also only tested at 3.9A, so things may be much different at higher of lower amps.
I am not kidding when I say that this does not look green to me. It really doesn't. The bare LED doesn't, and I put a lens up to it and shined it on the wall and it doesn't look green. I asked Lisa and a friend that is here today if they thought one looked different than the other (not knowing which one was the new or old) and they couldn't really tell.
Here is a macro shot of the new and old (note that there is some perspective shift):
New
Old
Even despite the perspective shift (there is some) you can tell that the die size is slightly different.
So, for the small sample size (only 2 emitters from 2 batches):
- Green hulk dedome: BUSTED
- Higher vF: BUSTED
- Bigger die size: VERIFIED
Thanks for the test but I am not sure what you want to tell with it?
Are you actually want to tell that new XP-G2 is good in your opinion?
It seems like you put some kind of lacquer on dedomed emitters (that kind of treatment leads to bit cooler tint) Vinz(famos German modder) called that treatment “diamond treatment”.
So RMM… You claim that new XP-G2 is not green when de domed? I just wish to know if anyone here on BLF agrees with you on that?
IMHO it has greenest beam projection I ever seen In my life.