Sofirn is unfortunately unable to source XP-G3 deep red from their suppliers. They can only order XP-E2 deep red. I asked them to order some samples for testing with a C01S driver to make a prototype (checking tint quality, lm/W ratio, optimum current for low/high mode). As monochromatic light is difficult to measure in terms of lumens the output quality will also be "measured" by the human eye. I hope they can do two samples, each with a different LED, i.e. XP-E2 red 625nm and XP-E2 deep red 660nm and put them side by side for comparison. Maybe there's a slight chance to get both versions but that is pending further testing and Sofirn's decision in terms of the expected sales forecast.
Furthermore, they will probably use the standard red body color as "wine red" will be difficult to realize by their anodizing factory. Maybe it's also a matter of costs for a small-size batch like this one.
The XP-E2 660nm is a recent quality led and is optically more suitable than XP-G2 (bit more throwish) so it will not get much better if you insist on 660nm.
Thank you for the update Lux-P. Anyway, I’d be happy if this project just comes into life. XP-E2 red is not a bad LED too, so it’s a not deal-breaker for me, hope for others too.
Just have in the back of my head Clemen’s store and E21A red on 10mm mcpcb. A nice mod idea… Have one spare waiting to be fixed
Looking at LH351H datashhets (especially the V2 option), it seems to be a bit more effective than XP-E2, but probably not worth the hassle of sourcing it.
Anyway, I’ll be interested in 660 nm option if it happens.
I know the discussion is pretty far along, but since people were interested in the XP-G3 and Sofirn is leaning away from it, here’s a comparison of available deep/photo red options. It looks to me like djozz is right. The XP-E2 is not far behind the XP-G3 unless you need the higher power handling:
MFG
Series
PWL (1)
Output at 350mA
I (max)
Notes
Cree
XP-E2 PR
660nm
18 lm (2)
1.0A
Many sources
Cree
XP-G3 PR
660nm
21 lm (2)
1.5A
Digikey
Luminus
SST-10 DR
660nm
16l m (2)
1.5A
Mouser, Digikey
Samsung
LH351H DR
660nm
18 lm (2)
1.0A
Digikey
Nichia
R E17A
650nm
31 lm
0.7A
Virence, phosphor converted
(1) Peak wavelength (PWL) is power based. Dominant wavelength (DWL) is perception/luminosity based. Different sources may use different specs. 660nm PWL ~= 645nm DWL (2) Datasheet gives output in mW. Lumens estimated using a luminous efficacy table at peak wavelength.
Note that the Nichia E17A’s output is a bit misleading. Just having a peak at 650nm instead of 660nm is as much as a 70% advantage in lumens, and being phosphor converted also helps, but shifts the DWL down to 618nm, compared to 645nm for the others in the list. It also has the lowest maximum drive current, and the highest forward voltage. The phosphor converted red, however, should look more like a filtered incandescent than the typical monochromatic LED.
Luminus has an SST-20 Deep red that is rated for 3.0A maximum, but I don’t see it in stock anywhere.
Thanks a lot for this overview. Unfortunately, Sofirn seems to be unable to source LEDs from other suppliers like Digikey, Cutter, Arrow or Mouser. They also told me that Nichia LEDs are hard to get. They have now applied for their supplier's MOQ of 1000 pcs of XP-E2 deep red. I told them this is a very nice LED for more than just the C01R (e.g. suitable for a Wurkkos WK30 update or for using it as auxiliary LED in addition to a powerful white beam flashlight like Nitecore P18). Moreover, I do have faith in people's growing interest once we get to see some first pictures of the C01R's beam. :-)
Since high currents do not apply to the C01R I am not so worried about 1 or 2 lumens less than XP-G3 deep red. With its small die size under the C01S optic I expect the XP-E2 deep red to put out a nice, useful but not too intense hotspot of red light.
By the way, we are now at 94.5% fulfillment rate to reach Sofirn's MPQ. :-)
I do not agree. 660nm is so invisible to the human eye that you really must adore that colour to justify spending so much energy to get the illumination at an acceptable level. So for aux leds and additional illumination they are a bit obscure in my opinion.
Edit: I have used 645nm 0603 leds as aux leds that look as good as the same tint of red as 660nm while producing light twice as efficient.
Au contraire, with all due respect. ;-) In my opinion deep red is most appropriate for lowlight scenarios where someone needs close-range illumination without causing too much attention (= without being seen). For instance, this could be very useful for reading a map in full darkness without giving up the human eye's adaptation to darkness. It could also make sense for having light in your tent while camping without attracting insects (that are less or even not able to perceive this wavelength).
Policemen or soldiers may appreciate the fact that 660nm is less perceptible by someone standing nearby. So, with regard to a tactical flashlight that Sofirn may want to develop, an auxiliary red light with deep red tint might be the right thing. It's not about high lumens here. Most red (auxiliary) lights are way too bright to be useful without giving up night vision capability, e.g. Wurkkos WK30 or Nitecore P18. A sublumen or low-lumen red light would be much more reasonable.
After all, I think we can agree to disagree. But maybe we only have two different perspectives how useful deep red light can be. :-)
I wrote down the argumentation because there is a common belief that deeper red is better no matter what (some people even want 720nm what we almost do not see at all), while in fact there is a strong trade-off in visibility. Any good reason to precisely want 660nm over 625nm despite the lower visibilty is ok to me though.
The deep red (660 nm) light would be far more useful to me. So if Sofirn produces these, then then I would like two more (for a total of 4). If only the 625 nm orange-red light is produced, then I will purchase only two. Thanks!
Lux-Perpetua, thanks for project-managing this for us! You are the man!
I would not be surprised if the 645nm and 660nm emitters you saw were actually the same wavelength, reported two different ways.
Some manufacturers or retailers list the peak wavelength, which is where the maximum output occurs on an optical power basis.
Others list the dominant wavelength, which is where the maximum flux occur on a luminus basis - it is adjusted based on perceived brightness according to the luminosity function.
It appears a peak wavelength of 660nm from a monochromatic LED works out to a dominant wavelength of about 645nm.
On the other hand, I am skeptical I could tell a 660nm peak LED from a 645nm peak LED.
Oddly enough, Cree’s XP-E2 datasheet lists dominant wavelength for the red, but peak wavelength for the photo red. So the two are slightly closer in practical terms than the stated wavelengths indicate - the red has a stated dominant wavelength of 625nm, but from the spectral graph, the peak appears to be about 635nm.
The XP-G3 datasheet, in contrast, lists both peak and dominant together for the photo red.
