Finding true cyan LEDs

[netprince]

Just got my XREBLU-L1-R250-00K01CT-ND and powered one up, really blue. Not green-blue like the other images hanging around on CPF...

 

[netprince]

Just an update...

I got out my quark RGB, started comparing them in a dark room, and interestingly the XREBLU-L1-R250-00K01CT-ND does appear to be about midway between the green and blue on the quark.

The quark's blue LED is very 'dark blue' (not sure the right term). The XREBLU-L1-R250-00K01CT-ND by itself looks blue, but when comparing it to the quark, it does look much greener.

I tried taking a picture to show the difference, but my camera doesn't seem to pick up the colors correctly... strange...

 

[saabluster]

Well thanks for the suggestions guys. I am curious if anyone has seen a significant improvement by adding cyan/green. Personally I just don't think it is worth the extra energy draw. Other than balancing out some excess red in the beam I don't see what it is good for as I didn't see that much improvement in the way the leaves look outdoors. :shrug:

 

[2009Prius]

...

I tried taking a picture to show the difference, but my camera doesn't seem to pick up the colors correctly... strange...

You may try turning off the auto white balance or including a white background in the field of view.

 

[2009Prius]

Mixing blue and green LEDs will result in dull saturation and other color problems. I have fully mixable RGB PARs, and trust me, running up the blue and green yields a muddy 'cyan' at best.

...

Sounds like the spectra of at least one of the LEDs in the mixture is "dirty" to start with. Too bad...

 

[2009Prius]

Agreed. Mixing two colors does not make them combine into a spectral output in the middle. It's not the same. ...

Many different spectra may be perceived as the same color by the human eye, which has only 3 types of receptors (some women have 4 the rumor says).

 

[netprince]

Another possible use, treatment of SAD...

http://en.wikipedia.org/wiki/Seasonal_affective_disorder#Treatment

 

[thepaan]

Well I now have the emitter and it is more blue than cyan. At very low currents it is almost a perfect cyan but turns more blue as the current increases. The search continues.

Blue LEDs shift towards UV with an increase in current but shift towards IR with an increase in heat. The shift shouldn't be greater than a couple nm so I'm surprised you notice it at all. I realize a high temp also shortens the life of the product but you could try not using an overly adequate heatsink - lets say one that allowed the LED to get up to say 50 or 60*C. Just a stab in the dark here....

 

[IsaacHayes]

You used to be able to get true cyan cyan luxeons, but now they are all traffic green. You all are on the right path to go for a blue luxeon that is binned very high towards green. 485nm is perfect.

I have a lux3 driven at 1 amp and it's perfect. Took me MANY tries to get what I wanted... My ISP deleted my subfolders on my website, and I can't find the original mod pics. big time upset about that, man this sucks.

When driven harder, it will go bluer. If you let it get hot, it will go green. Mine is cooled well and ran at 1 amp (dual BB500's, on 2 NiMH C). I ran dual BB500's as 1 set at 1amp would exceed the input current limit of the board, since the input voltage is quite low. By having 2 they share the load and it works great.

 

[anuragwap]

I resurrected this old thread since I've found a cheap source of true cyan emitters. Well,not quite. But they are in 495-500nm (bin2 cyan) range. These are the so called 'Exotic LEDs' available through LEDGroupBuy.com. Comparing side by side to my (dull) bin3 cyan Lux-V, I am now sure that these are indeed 495-500nm dominant at 700mA.

The other reliable option is ordering 1000+ cyan rebels from Luxeonstar, which is silly. Nichia/Avago has refused to accept bin specific order for high powered cyan LEDs even if ordered in full reel.

Another option, not very well known is using very low efficiency 1W emitters (490nm peak) from Epitex, a Japanese LED chip manufacturer. These are available from Roithner Lasertechnik/Marubeni America at ridiculously high prices/MOQ.

There's a Hong-Kong based company called YesLed which seems to manufacture (maybe in the past, probably didn't update their catalog) 485-495nm led dice, but no emitters.

I made a shower-head arrangement of 5mm LEDs with 217 rank C (495-500nm) NSPE510DS LEDs, resulting severe back pain after soldering the leads! I overdrive the cluster to 2.5-3A(for short time) to get a decent (bin1 equivalent) cyan.

 

[blasterman]

Steve's LED's sells a 'Cool Blue' Rebel ES that specs between 470-480, but to my eyeballs is no less than 480nm. They are pretty green.

Thought about using them to build an insanely high CRI room light.

 

[anuragwap]

Ordered 1 cyan and 1 cool blue from StevesLED.

UPDATE: Bad luck, their cyan is ugly bin4, and their sky blue is too blue for me, only a bit lighter than the common blue Cree c503 leds. To make up for my loss, I decided to mix them by wiring in series. The result was a nicely saturated greenish blue. Maybe I should mix more cyan's with fewer sky blue's to get a balanced cyan.

 

[PhotonFanatic]

Cyan emitters--I've still got some of the best that Future ever made-although they were Lux Is.

But when driven at 750mA, they are very bright and just the most beautiful color cyan you could want.

Yeah, that drive current was suggested to me by someone at Future--he said that while the lifetime would drop, it still would be close to 5,000 hours.

 

[Canuke]

UPDATE: Bad luck, their cyan is ugly bin4, and their sky blue is too blue for me, only a bit lighter than the common blue Cree c503 leds. To make up for my loss, I decided to mix them by wiring in series. The result was a nicely saturated greenish blue. Maybe I should mix more cyan's with fewer sky blue's to get a balanced cyan.

That's the best way to get the hue you want: mix readily-available wavelengths that are as close to the target hue as possible.

You won't get good cyan by mixing "standard" RGB colors, where green is around 520nm and blue at 460nm. Because of the green, that "side" of the RGB triangle leaves a large "lobe" of the color space outside its gamut; the closest approximation is significantly closer to the central white spot, i.e. there's not a lot of saturation. Cyan is actually the one hue that sRGB is least capable of rendering to high saturation.

Morevoer, the human eye is most sensitive to hue changes in this part of the spectrum. Look at the CIE curve; the evenly spaced wavelength markers are more spread out in that area than anywhere else on the curve. This is why shifts in hue are so much more apparent with emitters in this part of the band versus anywhere else; you won't notice a +/-3nm shift in a 650nm red emitter, but you will on a 488nm one. I'll bet that this effect is why there can be significant subjective variations between two people looking at monochrome cyan lights over what color they think it is, exactly. If you are looking for the "perfect" hue, you can most inexpensively dial it in this way.

 

[HarryN]

I think what you want is what used to be called blue bin 6. Not sure if that is still correct or not, but in that 485 - 490 nm range IIRC. The challenge is that it is easy to shift 2nm just from warming up, and that is a big shift at that color point. Try some of Photon Fanatic's Lux Is, you might be surprised how great they are.

 

[saabluster]

Thanks for the input in this thread all. I ended up finding what I wanted though. Avago has a cyan LED that I tested in a spectroradiometer. They were a beautiful cyan color. Here is the raw data.

Yellow marks the wavelength peak. Part #ASMT-JC11-NST01

400	1.142359E-012
401	1.305785E-012
402	1.491367E-012
403	1.707347E-012
404	1.949249E-012
405	2.21931E-012
406	2.515304E-012
407	2.832124E-012
408	3.158338E-012
409	3.538965E-012
410	3.903881E-012
411	4.268018E-012
412	4.641073E-012
413	4.998526E-012
414	5.348253E-012
415	5.676182E-012
416	5.996254E-012
417	6.30795E-012
418	6.608934E-012
419	6.905123E-012
420	7.189547E-012
421	7.478814E-012
422	7.760703E-012
423	8.050214E-012
424	8.338784E-012
425	8.63379E-012
426	8.935095E-012
427	9.245154E-012
428	9.567313E-012
429	9.907047E-012
430	1.026704E-011
431	1.065986E-011
432	1.106232E-011
433	1.151091E-011
434	1.197769E-011
435	1.250794E-011
436	1.311093E-011
437	1.37523E-011
438	1.447858E-011
439	1.528843E-011
440	1.621297E-011
441	1.724605E-011
442	1.840214E-011
443	1.972687E-011
444	2.12604E-011
445	2.296123E-011
446	2.490797E-011
447	2.714644E-011
448	2.971325E-011
449	3.257676E-011
450	3.580618E-011
451	3.959661E-011
452	4.377275E-011
453	4.849156E-011
454	5.376556E-011
455	6.002702E-011
456	6.704541E-011
457	7.469629E-011
458	8.350367E-011
459	9.34287E-011
460	1.046154E-010
461	1.17734E-010
462	1.321538E-010
463	1.482368E-010
464	1.662912E-010
465	1.870685E-010
466	2.100172E-010
467	2.352093E-010
468	2.633782E-010
469	2.936946E-010
470	3.270859E-010
471	3.650551E-010
472	4.065765E-010
473	4.536724E-010
474	5.087646E-010
475	5.700618E-010
476	6.385797E-010
477	7.18277E-010
478	8.056805E-010
479	8.982055E-010
480	0.000000001
481	1.108051E-009
482	1.218877E-009
483	1.339101E-009
484	1.455233E-009
485	1.57962E-009
486	1.702691E-009
487	1.827023E-009
488	1.944522E-009
489	2.056536E-009
490	2.163176E-009
491	2.252263E-009
492	2.331879E-009
493	2.393029E-009
494	2.438189E-009
495	2.46402E-009
496	2.470807E-009
497	2.459249E-009
498	2.430368E-009
499	2.385704E-009
500	2.327053E-009
501	2.25551E-009
502	2.177913E-009
503	2.091755E-009
504	0.000000002
505	1.906136E-009
506	1.808378E-009
507	1.713306E-009
508	1.620935E-009
509	1.528111E-009
510	1.440092E-009
511	1.353841E-009
512	1.270531E-009
513	1.195086E-009
514	1.122061E-009
515	1.053664E-009
516	0.000000001
517	9.248806E-010
518	8.658131E-010
519	8.133777E-010
520	7.606843E-010
521	7.120525E-010
522	6.638793E-010
523	6.182568E-010
524	5.756395E-010
525	5.355418E-010
526	4.973734E-010
527	4.613203E-010
528	4.286159E-010
529	3.969579E-010
530	3.676203E-010
531	3.404806E-010
532	3.15927E-010
533	2.947597E-010
534	2.7322E-010
535	2.534005E-010
536	2.355758E-010
537	2.192256E-010
538	2.037059E-010
539	1.892112E-010
540	1.756544E-010
541	1.629717E-010
542	1.51238E-010
543	1.399529E-010
544	1.293977E-010
545	1.194971E-010
546	1.101403E-010
547	1.018992E-010
548	9.401337E-011
549	8.669414E-011
550	8.009427E-011
551	7.404531E-011
552	6.857022E-011
553	6.343328E-011
554	5.879982E-011
555	5.444538E-011
556	5.04516E-011
557	4.67068E-011
558	4.330378E-011
559	4.011693E-011
560	3.714329E-011
561	3.437823E-011
562	3.173883E-011
563	2.940898E-011
564	2.723758E-011
565	2.521905E-011
566	2.336279E-011
567	2.16593E-011
568	2.002472E-011
569	1.857551E-011
570	1.728397E-011
571	1.606538E-011
572	1.492689E-011
573	1.38695E-011
574	1.28975E-011
575	1.198574E-011
576	1.114145E-011
577	1.033596E-011
578	9.609101E-012
579	8.931806E-012
580	8.308541E-012
581	7.740564E-012
582	7.210307E-012
583	6.707014E-012
584	6.256915E-012
585	5.843098E-012
586	5.457412E-012
587	5.09004E-012
588	4.77324E-012
589	4.474127E-012
590	4.193354E-012
591	3.930163E-012
592	3.693058E-012
593	3.468128E-012
594	3.25378E-012
595	3.059763E-012
596	2.873207E-012
597	2.706775E-012
598	2.546376E-012
599	2.4019E-012
600	2.264691E-012

 

[BVH]

Are these your "To Die For" Cyans? I bought a bunch from you waaayyyyy back when and still have them in my parts bins. I had one put in a Fenix pocket light and I have 4 in Mac's 4-Cyan, green cut down mag beauty! Eight power levels up to 1 Amp. Beautiful Cyan color!

http://www.candlepowerforums.com/vb/showthread.php?146752-Quad-Cyan-*SOLD*&highlight=quad+cyan

Cyan emitters--I've still got some of the best that Future ever made-although they were Lux Is.

But when driven at 750mA, they are very bright and just the most beautiful color cyan you could want.

Yeah, that drive current was suggested to me by someone at Future--he said that while the lifetime would drop, it still would be close to 5,000 hours.

 

[anuragwap]

Photonfanatic has some Q2H cyan left, no bin1 cyan or bin6 blue. And those 496nm peak Avago ASMT-JC11-NST01 LEDs should be about 499nm dominant, i.e. still no good

 

[Canuke]

Are these your "To Die For" Cyans? I bought a bunch from you waaayyyyy back when and still have them in my parts bins. I had one put in a Fenix pocket light and I have 4 in Mac's 4-Cyan, green cut down mag beauty! Eight power levels up to 1 Amp. Beautiful Cyan color!

http://www.candlepowerforums.com/vb/showthread.php?146752-Quad-Cyan-*SOLD*&highlight=quad+cyan

I have 5 of these, from Photonfanatic also. I've never seen this color anywhere but a 488nm laser, so I'm pretty sure that's where the peak is, within +/-2nm. I intended them for a clear PC case, with the intention of texturing the clear panels and lighting up the inside "ice blue" with these. I never did build it, as clear cases went out of style around then and I never solved how to get the texturing I wanted.

I did install one of the "bin 6's inside a cheap host, its color is almost perfectly matched to the Lux V "eye cryin' cyan" Brinkmann I have.

 

[SemiMan]

I resurrected this old thread since I've found a cheap source of true cyan emitters. Well,not quite. But they are in 495-500nm (bin2 cyan) range. These are the so called 'Exotic LEDs' available through LEDGroupBuy.com. Please do not misunderstand me, I am NOT advertising for them, I assume these are cheap Chinese LEDs with low junction temperature, poor lumen maintenance and are ESD prone, inefficient. But if you want bin2 cyan in high power, its the only sensible way to go. I had emailed the shop owner to test the LED color before ordering. Comparing side by side to my (dull) bin3 cyan Lux-V, I am now sure that these are indeed 495-500nm dominant at 700mA.

http://www.flickr.com/photos/66231005@N05/7754713982


The other reliable option is ordering 1000+ cyan rebels from Luxeonstar, which is silly. Nichia/Avago has refused to accept bin specific order for high powered cyan LEDs even if ordered in full reel.

Another option, not very well known is using very low efficiency 1W emitters (490nm peak) from Epitex, a Japanese LED chip manufacturer. These are available from Roithner Lasertechnik/Marubeni America at ridiculously high prices/MOQ.

There's a Hong-Kong based company called YesLed which seems to manufacture (maybe in the past, probably didn't update their catalog) 485-495nm led dice, but no emitters.

I made a shower-head arrangement of 5mm LEDs with 217 rank C (495-500nm) NSPE510DS LEDs, resulting severe back pain after soldering the leads! I overdrive the cluster to 2.5-3A(for short time) to get a decent cyan.

http://www.flickr.com/photos/66231005@N05/7754792738

Have you contacted Lumileds/Future Electronics? If you are buying a whole reel, they are likely able to find you a specific reel. If you want a bunch of reels, that may be an issue. There may be some premium of course. It used to be hard to get specific bins as all the good cyan bins went into traffic lights. I am not sure if that is still the case.