Throwy 4000K would be nice, considering SFT40 4000K is practically nonexistent.
3 Thanks
I’m looking to purchase my first skilhunt lights. Really want an M200 519A V4 + the new F35R.
Will you still keep 519a as an option when the new emitter releases? Also will a V5 be released with the F35R or will it still be a V4?
Thanks for your interest!
Right now, the new F35R emitter, it’s not yet decided which models will get it.
1 Thank
I, for one would really like to see it in the M150 and M200.
I’m also curious about how well it would work in the EC500.
1 Thank
Any info of models you choose for F35R emmiter? Please, tell me EC150 is one of them…
1 Thank
I would also like to know the answer to this, if skilhunt sees this.
1 Thank
A thought: it seems there is still no good cyan phosphor with λex = 470-480 nm available, especially not in a broader FWHM range. There is BaSi2O2N2:Eu2+ or NKLSO:Eu2+ available, but both have it’s peak at around 490 nm which is too high for classic blue pumps to fill the gap completely.
All known phosphors for 470-480 nm are excited in the UV range (350-400 nm) which is way too low since a (thermally more sensitive) UV chip would be needed. Also the temperature stability / saturation at high radiance could be an issue, but I have no clear information found on this yet.
This is maybe the reason why no high power LED with completely filled cyan gap is available. There are some emitters around which has a partially filled gap (Bridgelux Thrive, I have some of these so I will test them in the next weeks) but they also have some sort of cyan gap.
1 Thank
Many interesting thoughts, thank you for sharing this!
I think this might not be a huge problem:
- Having an additional 490nm peak is already a very substantial improvement over the existing phosphor mixes. I don’t know what component phosphors are used in emitters like 519A (do you have this info for some blue-pumped high CRI LED?), but it looks like a secondary peak (after the 450nm blue) doesn’t occur until 510nm or after.
- One might not need emission centered at the dip in order to fill the dip. Both the 450nm pump and 490nm phosphor will have heavy tails surrounding the peak wavelength, and there is a chance that the overlap of the 450nm’s right tail and 490nm’s left tail might suffice to fill the gap. Here’s a cartoonish, exaggerated illustration of what could happen when you add the two spectra:
Do you happen to know what phosphor is used in Nichia’s Optisolis emitters, or Seoul’s SunLike 3030 emitters? If I recall correctly, they are both pumped with 420nm.
So you are saying that the weakness comes from the phosphor unable to handle high temperature or high power density? This seems sensible, I think the technology of UV emitters has matured enough that the UV die itself should not be the bottleneck.
If someone can get hands on a high power 420nm emitter and transplant some Optisolis/Sunlike phosphor on it, it would be interesting to see how much power the phosphor can handle. Or one can try to overdrive an Optisolis/SunLike and see what fails first: the phosphor or the die. I have some low-power 3535 SunLikes and will try this and see what happens. I suspect the die will fail first because the phosphor looks diluted by some sealing substance.
For all visual purposes, I would consider the Thrive to have a completely filled cyan region–the difference is astronomical compared to other blue-pumped LEDs!
So LuAG phosphor could be enough…
I tested this ealier with some of my reference color lights (which includes a cyan variant) and there was still a relatively huge gap visible. Maybe I can post some images tomorrow.
Good question, they have a 450 nm peak, presumably created by phosphor. There are some phosphor types like SMSO:Eu2+ or SHSO and even BAM out there (the latter is used commercially) but they have all excitation peaks in the UV-A range. There is BSOC:Eu2+ but this is a novel phosphor type which is not used commercially so far I know.
Could be. I am not sure. It seems the junction temp for UV chips (365 nm or even below) is very low, for Nichia UV emitters at 100 °C which would mean they cannot operate under ambient temp of more than 50-60 °C, depending on their drive current and the heatsink/thermal resistance. I think this is unacceptable for a modern high power LED in the 10-20 W range.
I think there is a reason why there are currently no UV-pumped high-power LEDs in the 10 W range or above. The phosphor itself does not seem to be the problem to me, as there should be good substitutes with excitation wavelengths in the UV-A range, especially with newer research in novel rare-earth free phosphors, which might be a huge thing in the future. I really think it’s the chips. Anything below 400 nm still seems to have too low Tj to be used effectively in a normal environment with corresponding temperatures.
We will see how they perform, especially since they use a classic blue pump instead of a violet one.
I saw an interesting difference between the phosphor layer of Sunlike 3030 and Thrive LEDs:
The phosphor of the Thrive emits way more red light under 365 nm UV than the Sunlike, even in the warm CCTs. This is really interesting and raises the question on how much phosphor Bridgelux had to use to achieve this spectrum in comparison to the violet-pumped Sunlike.
3 Thanks
That would be extremely illuminating!
If this were the issue, shouldn’t we already see it in UVA emitters? There is a huge gap in power handling between even modest UVA emitters and the highest-power single-die violet-pumped emitters. Some recent UV emitters (e.g., SST08-UV-H) have very impressive current density above 1A/mm^2.
The part the puzzles me is that the violet-pumped emitters always come with tiny little dies (on the order of 0.1mm^2), that is the reason why they cannot handle much power. Some emitters like Optisolis even drive two in parallel. How hard could it be to just make a single bigger die?
I am inclined to attribute this difference to phosphor composition rather than thickness; in particular, the fluorescence spectrum at 365nm may not be representative of the fluorescence spectrum at the pump wavelength.
One interesting example is LH351D 5000K, which appears yellow-green under 365nm unlike most high CRI (and even low CRI) LEDs of this CCT, which are deep orange.
I ran some tests.
You can fill the gap with Cyan phosphor (λem = 490 nm). But then the duv goes massively up so you have to use red phosphor to counteract this and lower the duv again. This in turn reduces efficacy.
It seems the cyan phosphor is really sensitive to radiance or temperature. There is visible tint shift with rising current.
There is the XE-G violet with 410 nm peak, which is rated to 150 °C Tj and 3 Amps. Most noticable difference is the way higher Vf (3.2 V at 1 Amp vs. 2.9 V for white blue-pumped LED).
Could imagine it is expensive to manufacture violet dies and/or yield is relatively low. Together with maybe way more expensive phosphor for excitation in this wavelength range the manufacturers simply don’t want to produce these since blue-pumped high CRI emitters are good enough for 99.9 % of all use cases.
1 Thank
Thanks for running the test! Since adding blue drops duv and adding green increases duv, I expected adding cyan to more or less leave duv unchanged, but of course that is a vague and unquantified heuristic, and 490nm is probably too close to green.
I see about the reduction in efficacy. I personally wouldn’t mind at all in exchange for much better light quality, but it would fare poorly on the market where people just care about output numbers.
Now that is a more fundamental problem…
This is likely the true reason why nobody has the incentive to produce high-powered violet-pumped LEDs. The difference between 95 and 99 CRI probably means nothing to most people, but the lumen count certainly does.
1 Thank