Cree changed production process for XP-G2 emitters!

Sweet! If you get a chance to try this, and it works well, you should add it to the Perfect Dedome thread!

I have posted in other threads that I use this KleanStrip. It’s a paste that I set on top of the led dome. I just used it to successfully de-dome a Nichia 219C. The key is to set a “blob” on top for 20 minutes or so while occasionally swirling the gel around the led. After this time I remove the gel by blowing it off with compressed air. It usually takes a few rounds of this to completely dissolve the dome. I have never had a bad tint shift either, just makes it a bit warmer.

As a side note I was only able to get 140K @ 1M corrected at 10M, from the 219C in my C8. When I put in an “old-style” XP-G2 S4 I got 185K.
I think my numbers could have been better but I may have been “over driving” the 219C with spring bypass on both ends, 30Q cell and MTN FET driver. Thoughts?

I don’t think you’re alone with the 219c, others have had lower throw numbers for whatever reason. It is very possible it was on the downward end of it’s intensity with that FET/30Q setup as well.

I’m very interested in your kleen strip stuff. I just might go get some tonight.

Yeah, here’s a thought. Are you up for some testing on the newer LEDs? Several around here would love to know what you can get from the new XP-G2 and the XP-G3 with your same set-up for comparison to those other two you mentioned. Stand-alone measurements are nice, but comparisons are better. :wink:

Edit: Oh Yeah! My local Home Depot has that Kleanstrip Adhesive remover! :smiley:

219c wont work folks, its just its core is way bigger than S4- same goes for the new S5

meanwhile S4 2C hit 5.2 amps here( on of them, the other 4.5)….202k mcds in a C8 host….pretty awesom….will be awesom too if Cutter responds my PMs…eventualy

P.S
that S4 2C went staight to my .223 hunting M2X and it topped 370k mcds with a fresh VTC5A( it was 340k before that, S2 2b at 5.5amps)
The tint is gorgeous for hunting realy

Sounds good to me. I could use the same Thorfire C8 with AR glass and FET driver for all testing. I have a good dark room that I have measured out 10 meters. The only issue is my funds are a little tight right now. I currently only have a couple old style G2 S4s and G2 S2s. If anyone has a XPG3 or other emitter you want me to test I would gladly test them for you. :wink:

Edit: Just saw Mitko’s post so probably not worth the time. Offer still stands though LOL

Those are awesome numbers! I think that is the first time I have seen anyone claim 200k+ from a C8 host.

I have a theory about what happened to the die size of the XP-G2 S2. Is it possible that the new G2 S4 2B shares the same die as one of the new XP-G3s? The XP-G3s get there additional lumens from the phosphor dome but Cree uses the same die? And it could be possible that other tints such as 2C don’t share dies with new XP-G3s and therefore will continue to be manufactured with smaller dies? Speculation and wishful thinking I’m sure.

I should post about the XP-G3 emitter test first before posting about my slicing experiment, but this evening I had enough time to finish the throw test and way too little time for the emitter test write-up, so here you go:

The sliced led looks like below at 0.5mA, this looks similar to a sliced 3rd gen Oslon Square, or a Luxeon Q (and probably Cree XT-E looks like this too but I never had one of those). All these leds have a ‘poured over’ phosfor layer and do not throw very well. I sliced the silicon very close to the phosfor, in fact a few little flakes of phosfor were cut off with the last slice. The finish is not completely flat but I do not see how a perfectly flat shiny finish would improve throw immensely.

I made a direct drive pill for the Brinyte 158, I have another pill for it with a XP-G2 S4 2B that did 322 kcd. In the picture is on top the XP-G3 S5 3* pill, in the middle a XP-L Hi for comparison, bottom is the XP-G2 S4 2B pill. The XP-G3 die is slightly larger than the XP-G2 die, by (rough measurement) 13%

In this light the XP-G3 S5 3* draws 7.3A from a fresh 30Q battery, which is very close to its maximum output (99.2%)

The throw I get in the Brinyte with a Samsung 30Q cell, measured at 7 meter, 30 seconds after switch-on is 206 kcd, so way short of the XP-G2 S4 2B. I re-measured the throw of the old G2 S4 2B pill and it still throws well over 300 kcd with the partly drained 30Q from the first test.

I know that BLF member MEM claims some dedome magic, I hope his dedomed leds perform better than this one.

Thanks for the testing djozz.

:weary:

7amps though…. can’t wait to see the full writeup

Yep, the voltage is not as low as the Nichia 219C, but pretty close! And the led is tough as nails!

You should try cutting away the extra phosphor, unless you had plans to use that LED… :smiley:

It looks like the unmounted S4 emitters at cutter are now on back order.

I tried that once with a Luxeon Q led and it did not improve throw :frowning:

Remember those real inexpensive XPG2’s Fastech sold about 3 years ago? Are those actually desirable in comparison to these new versions?

Except, that photo is not dedomed. That’s sliced, like a HI version. I completely dedomed the XP-G3—bare. The S5 3A dedomed looks just like a dedomed Nichia 219C in tint, also. Probably the two closest tint LEDs I’ve seen, dedomed.

Luminarium. I found something extremely perplexing about this entire ordeal over on my end. Now, maybe I am wrong. Maybe I compared 2 other emitters, that are not the same as your “big” and “small”. But I did compare two “new” and “old” Cree XP-G2s to find this interesting thing. What I found, is that my new Cree XP-G2 indeed had a die that looked larger than my old XP-G2. That is, until I measured them to find that they were the same size. What was actually different about them was NOT the die size, it was the anode/cathode insulator line—a package change—that had the insulator line closer to the die perimeter, leading one to believe at first glance that the die was much larger. But it wasn’t!

Has anyone confirmed this, 100%?

I too noticed the color difference just like was stated; new had more of a dull yellow phosphor color, old had more of a bright fluorescent-yellow phosphor; dedomed. But my two indeed only varied in that line around the die, not the die sizes. Another person has also said he noticed no intensity difference between them. I am starting to wonder if some of this is actually human error/perception-based, and not a die which is larger. I bet that any person would say the “new” die is larger 10 times out of 10 when I put them side-by-side on the table because of this package line change, but in reality with my two, no.

I really don’t think anyone should ever think to go after the person that sold them these LEDs. The LEDs are only “bad” in your mind, as you are the outcast user of them. I read some of that earlier about pulling money back from the supplier and I couldn’t believe someone would want to go after a supplier of something that Cree handed to them under the same specs they had always been dealing with. You can’t expect a supplier to be pulling out his caliper to gauge each LED die size, after dedoming one! That’s just not how this stuff works; only inside of our extremely small hobby-bubble on a corner of the internet does it work in die-sizes and high-kcd throw. The industry works in those very specs you see from Cree, but no where does Cree guarantee a die size, or certainly a lux value that cannot be reached. :wink:

MEM,

Don’t get me wrong… But you did not read a word from this thread. Which person stated that there is no performance difference between them?

New production process XP-G2 has:

- slightly bigger dice or die,

- the number of dots on the die have increased, the phosfor is rougher.

- the tint of the phosfor is a bit more orange with small red bits in it which are absent in the old die

  • when de domed it brings lower performance in all flashlight configuration you can imagine

Read first post of this thread you got plenty of info there

But that really don’t concerns me any more… You can think that you got bingo with new production type XP-G2(you are one of rare persons who think so)


What concerns me now is that new XP-G3 also fails to bring performance of good old XP-G2 S4 2B

Djozz that test is more than adequate to me. Not even pure de dome(without slicing) can’t improve it that much.

Thanks Djozz :slight_smile:

Sigh, sigh, sigh… Id also take 100x old batch xp-g2s…

I have 9 XP-G2 S4 2B laying around for a project. I ordered them a year ago from intl outdoor.
Can someone please confirm they are the good old production typ ones.
(Click image for full size)

From the post of luminarium , i think they are the good old s4 2b :wink:
Really lucky , I’m searching for 2 without luck :wink: