Trying to find a blue-green wavelength LED

[Anders Hoveland]

Hi to this forum.

I am trying to find a blue-green wavelength high-power (at least 1 watt) LED emitter. I am looking for anything between 480-490nm. I am having a very difficult time finding anything in this spectral region.

To make matters more difficult, there are countless LED emitters being sold that are "mislabeled". For example, most of the "ice blue" LED products that are labeled as "485nm" are actually more like 470nm with a green phosphor. So they are more like an "effective 485nm" color LED, but in reality there is a big dip in the spectral output right in that region. Mouser sells Cree blue emitters sold as 485nm, but these are actually 465-485nm, with the peak at about 473nm. I have been looking for 490nm blue LED also, but most of them labeled "490nm" are actually 490-495nm, with the peak much closer to 495nm, so they are much more of a minty-green color than blue or real turquoise.

In this particular region of the spectrum, a small difference in nanometer wavelength makes a very significant difference in color. 495nm is definitely green, while a true 490nm emission would be pretty blue. Of course, LED emitters in practice do not emit at a single discrete wavelength, and while their bandwidth tends to be rather narrow, in this region of the spectrum "narrow" starts to get more relative.

I also looked at the 490nm blue LED source at Thor Labs but it was rather expensive and I will need several of these things.

I am also open to the possibility of combining a blue and cyan LED to get the desired spectral peak. However, to do this the cyan LED would need to be very close to 490nm and the blue LED would need to be very close to 480nm for there to be significant overlap. I have tried a 470nm and 495nm LED and this did not work. There was only very minimal spectral overlap. Using a Cree XPE cool blue emitter instead of the 470nm worked better, but I could still tell there was a 30-40% dip in the spectral output at ~485nm.

I know there are many knowledgeable people in this forum. Is anyone aware of an LED emitter that would meet my needs? What is the "greenest" tinted blue LED emitter that they make?

 

[SemiMan]

Nichia:

http://www.nichia.co.jp/specification/products/led/NCSE119A-E.pdf

Closest I am aware of. It more like 500nm, but not aware of anything current mass marketed at 490nm.

Problem is as it gets warm it will shift to longer wavelengths.

 

[Anders Hoveland]

Seems like this is an elusive wavelength range, trying to find an LED manufacturer who makes light blue or turquoise color emitters in high power.
Why the gap? It's easy to find a 470nm LED or a 495-505nm LED, and they make LEDs that cover almost every other area of the spectrum.

I know color can be very subjective, but I just do not really find 495nm to be bluish at all. It is not really turquoise, seems more like a minty pale green to me.

 

[Anders Hoveland]

I have read on this forum (and other forums online) that a Luxeon blue bin 6 is more of a cyan blue. Many people were complaining that some of their blue LEDs were too greenish, but this is exactly what I am looking for. Unfortunately, all the references to this I can find are at least 5 years old. Maybe they no longer offer these greenish-blue bins? Does anyone know where I could get one?

 

[Ono Nadagin]

XLamp XT-E Royal Blue

Color Dominant Wavelength Royal Blue 450 – 465(nm)

Minimum Radiant Flux @ 350 mA, 85°C 475 – 600(mW)

 

[anuragwap]

Anders, like you I've also tried those 'ice blue' chinese LEDs with green phosphor and have been disappointed. (Their color saturation is very low because of the mixing of blue and wideband green emission, they look more like a bluish white than cyan). I've also tried 490-495nm dominant LEDs from different manufacturers: Luxeon Bin1 cyan rebels (search 'true cyan' to find that thread) first, then some fat domed and some XP-E clone typed Chinese emitters with Epiled/Epistar chips. When run at 350mA, they are all roughly 493nm dominant, and 491nm peak. At 700mA-1A with good heatsinking, they might shift to 490nm dominant, 488nm peak at best.

480-490nm dominant Luxeons have long been discontinued acc. to the latest Luxeon rebel datasheets, although Cree datasheets for XP-E and other color LEDs still mention about the blue B6 bin(480-485nm dominant). But this again is very hard to get (try your luck at Cutter/Digikey/Mouser/LEDSupply etc.). Most of the cool blue LEDs sold in market today are 475-480nm dominant.

 

[Anders Hoveland]

Anders, like you I've also tried those 'ice blue' chinese LEDs with green phosphor and have been disappointed. (Their color saturation is very low because of the mixing of blue and wideband green emission, they look more like a bluish white than cyan). I've also tried 490-495nm dominant LEDs from different manufacturers: Luxeon Bin1 cyan rebels (search 'true cyan' to find that thread) first, then some fat domed and some XP-E clone typed Chinese emitters with Epiled/Epistar chips. When run at 350mA, they are all roughly 493nm dominant, and 491nm peak. At 700mA-1A with good heatsinking, they might shift to 490nm dominant, 488nm peak at best.

Yeah, do not waste your time. Anything made in China that says "480-490nm" or "485nm" is crap. I just received an order for some 10W emitters from China sold as "490nm". The picture of the light color when I placed the order was very misleading. It is actually just a regular 495nm with a little blue sticky plastic covering the LED! And when powered, it only looks a little more turquoise than regular 495nm, it is still very green.
Chinese sellers are very deceptive. If it does not have phosphor over it, chances are it will contain a double chip with a blue and green wavelength LED combined.
I have made several purchases from China, desperately looking for anything that might come close to a greenish-blue wavelength.

480-490nm dominant Luxeons have long been discontinued acc. to the latest Luxeon rebel datasheets, although Cree datasheets for XP-E and other color LEDs still mention about the blue B6 bin(480-485nm dominant). But this again is very hard to get (try your luck at Cutter/Digikey/Mouser/LEDSupply etc.).

I got the "485nm" emitters from Mouser. They are rated for 3 watts and I am running them on only 1 watt, so that should increase the wavelength about +3nm. It is difficult to tell, but they do seem to be in the 480-485nm range.
It is a nice soft blue, but it is definitely pure blue, not the slightest bit cyan.

I have seen some beautiful pictures of 488nm DPPS laser diodes, but unfortunately they are far to expensive to be using for general lighting purposes.

It is so frustrating that they do not seem to make any bluish cyan LEDs.

When run at 350mA, they are all roughly 493nm dominant, and 491nm peak. At 700mA-1A with good heatsinking, they might shift to 490nm dominant, 488nm peak at best.

Bluish-cyan (485-490 ?) is such a beautiful color, but increase the wavelength just +5nm and it become a murky white-mint-greenish. I find this color ugly and annoying for some reason. So frustrating...


I found this: http://www.imaginarycolours.com/Mer...de=QDG12Aqua&Category_Code=QD&Product_Count=4
Aqua color LED Christmas lights based on recent quantum dot technology. Unfortunately they no longer seem to be sold. The product was called "Dotstrand" from Evident Technologies.
http://www.jimonlight.com/2008/10/17/crazy-colors-in-christmas-leds-this-year/

 

[Anders Hoveland]

I found this little online application:
http://users16.jabry.com/lsrtools/WaveToRGB/

You can correlate any wavelength value to an exact color.

 

[brickbat]

Why would that be exact? seems more like a very rough approximation, based on whatever monitor you happen to be using...

 

[Anders Hoveland]

Table 12. Dominant Wavelength bin structure for blue LUXEON, LUXEON K2, LUXEON III and LUXEON V emitters and arrays

Bin 1	460	465nm
Bin 2	465	470nm
Bin 3	470	475nm
Bin 4	475	480nm
Bin 5	480	485nm
Bin 6	485	490nm

tried looking for part number LXK2-PB14-N00, but they seem to be discontinued, having a difficult time finding a distributor that still has them in stock. also tried looking for part number LXHL-PB01 but specs say it emits only 16 lumens.

 

[Xe54]

Here's a crazy idea, if you can tolerate pulsed/PWM light: Take a Lumileds Rebel Cyan LED, and drive it at a low duty cycle (5-10%) with about 5-7 amps. I'm not sure of exactly how much wavelength shift this will cause (I could measure it at work, but my boss is hounding me about other stuff, so I haven't much play time). But my experience with the green LEDs is that they shift to a solidly cyan color (+10 to 15nm!) when driven the way I drive them for research purposes (up to 20 or 30A for 1 mm^2 chips, for several microsecond pulses). Doing this at high enough frequency to completely eliminate perceivable flicker is possible, but would require some custom electronics beyond the typical LED driver ICs.

 

[anuragwap]

Thanks for the suggestion! But 5-7 amp to that small die? How much V_f are we talking about at that drive level, supposing junction temp is kept under control by setting low enough duty and good heatsinking? (And I feel 5-10% duty will be too much)

I love my bin1 cyan rebels too much to try this, since I got them after wasting $$$! But I can always try the Chinese equivalent cyan LEDs as guineapigs, interfaced with mosfet/pic at a pwm of few khz, while they'll complain to me about the torture by making audible whines at these high current levels! (I've direct-driven 5mm Nichia NSPE510DS LEDs from 6V Lantern battery for a couple of seconds,the bond wires didn't melt, my slow Fluke DMM set to max mode registered 500mA, and the color was extremely pleasant!)

 

[Anders Hoveland]

What if one used a much higher current (x10) power supply with a pulse-width modulation dimmer? These dimmers are common and relatively cheaply available.
Would that be enough to shift the wavelength 5-6nm shorter?

 

[anuragwap]

Anders, it can be done just to see the beautiful argon laser-like cyan color, but keep in mind that efficiency will be horrible due to droop of green/cyan LEDs. Just have a look at the Lux rebel datasheet, see how lumens eventually max out with increased current drive. (Blue/Royal blue are somewhat better in this respect) Besides, electromigration can occur at these higher current levels (even at low duty) which will eventually result LED failure.

 

[JoakimFlorence]

There has been a significant development.

Osram has released the OSLON PURE 1414 family of LED emitters.
(It was announced March 2024 and as of early November these are now commercially available)

Among these is the GC VJLPL1.13-KQKS-V2V3-1
It's a "cyan" color emitter offered in 3 bins. The V2 bin is the shortest wavelength (most blue in color), with a wavelength minimum of 485 nm and an (average) maximum of 490 nm. The technical specifications say the wavelength is centered at 488 nm.

1414 stands for square emitter dimensions of 1.4mm x 1.4mm.
It's rated maximum current is 1000mA.

1.4mm is very small and an unusual size, making it difficult to find a heat sink specifically sized and designed for mounting the emitter to.
It's also a very high power emitter relative to the tiny size, which one can assume makes optimal mounting to a heat sink even more important than usual.
Unfortunately, I think we can expect significant color shift (towards green) if this emitter is significantly underdriven at lower power level.

Looking at the indicated spectral graph in the specifications, the majority of the wavelength coverage area and peak of the spike appears to be below 490 nm, but around 35 percent of the wavelength area coverage appears to be above 490 nm.

typical luminous efficacy 89 lumens per Watt
rated maximum operating temperature 100 °C (lower than other high power emitters at 140 °C)
2.4-2.8v typical voltage 2.74
( 488 nm theoretically correlates to 2.5406 volts, on a quantum level)

 

[Monocrom]

1414 is currently available?
SWEET!!!

 

[JoakimFlorence]

1414 is currently available?
SWEET!!!

Good luck finding a PCB board or starboard that is compatible to mount it on, however.

(And at that extremely small emitter size, with 3 Watts of power, you want to make sure it is properly mounted and has an optimal heat sink connection to carry away the heat)

 

[DoctorMemory]

Hi to this forum.

I am trying to find a blue-green wavelength high-power (at least 1 watt) LED emitter. I am looking for anything between 480-490nm. I am having a very difficult time finding anything in this spectral region.

To make matters more difficult, there are countless LED emitters being sold that are "mislabeled". For example, most of the "ice blue" LED products that are labeled as "485nm" are actually more like 470nm with a green phosphor. So they are more like an "effective 485nm" color LED, but in reality there is a big dip in the spectral output right in that region. Mouser sells Cree blue emitters sold as 485nm, but these are actually 465-485nm, with the peak at about 473nm. I have been looking for 490nm blue LED also, but most of them labeled "490nm" are actually 490-495nm, with the peak much closer to 495nm, so they are much more of a minty-green color than blue or real turquoise.

In this particular region of the spectrum, a small difference in nanometer wavelength makes a very significant difference in color. 495nm is definitely green, while a true 490nm emission would be pretty blue. Of course, LED emitters in practice do not emit at a single discrete wavelength, and while their bandwidth tends to be rather narrow, in this region of the spectrum "narrow" starts to get more relative.

I also looked at the 490nm blue LED source at Thor Labs but it was rather expensive and I will need several of these things.

I am also open to the possibility of combining a blue and cyan LED to get the desired spectral peak. However, to do this the cyan LED would need to be very close to 490nm and the blue LED would need to be very close to 480nm for there to be significant overlap. I have tried a 470nm and 495nm LED and this did not work. There was only very minimal spectral overlap. Using a Cree XPE cool blue emitter instead of the 470nm worked better, but I could still tell there was a 30-40% dip in the spectral output at ~485nm.

I know there are many knowledgeable people in this forum. Is anyone aware of an LED emitter that would meet my needs? What is the "greenest" tinted blue LED emitter that they make?

Try digikey.com. Use their search function and get about any wavelength you want. Use first class mail for shipping, cheap small orders. I just bought some "traffic light green" 505 nm which are almost blue (as are the traffic lights).

 

[JoakimFlorence]

A Chinese-based manufacturer named LEDGUHON appears to be offering a 485-490nm cyan-blue wavelength (along with several other specific and less common blue wavelength options).

" 3535 SMD LED Blue Light 3W LED Chip Surface Mount LED Diode Super Bright Diode 485-490nm "

emitter dimensions 3.45x3.45mm
recommended maximum 700mA, peak 1000mA
the specifications give the wavelength as being centered at 487.5 nm (typical)

They appear to be selling through the Amazon site.
It seems like a little bit of a lower quality LED brand, but should still be suitable.
I have a suspicion though that the picture of the LEDs in the sales ad is not truly the exact real picture or appearance of what their emitters look like.