First-generation blue LED hue?

[jolshefsky]

In late 1992 to early 1993, I was working at Kodak for their research labs (for photographic film) so there was a push for measurable, controllable work lights, and LEDs seemed to fit the bill. One of the more senior researchers got hold of blue and RGB LEDs. As I remember it, the blue LED demo was a disk of about 50 LEDs that plugged into the mains. It was the first time I ever saw blue LEDs and I remember the color being a magical true blue, not like the deeper blue we see today. Does anyone know the wavelength of those first LEDs? Am I just mis-remembering the color?

[frozenfrogz]

The first blue LEDs used GaN. The inventors actually got awarded a Nobel price for their invention.
However, the very first blue LED was actually invented by RCA in the 1960s!

GaN LEDs are 450nm if I recall correctly.

[Kleinstein]

The first blue LEDs were SiC based and very low efficiency.

With modern GaN based LEDs the color depends on the exact type and can be somewhere in the 400-480 nm range.

I still remember the color of the very first blue laser pointer (around 1999) - a faint blue -violet at around 405 nm. Not very bright unless it hits something with a florescent dye.

[jolshefsky]

With modern GaN based LEDs the color depends on the exact type and can be somewhere in the 400-480 nm range.

I'm looking for ones at the longer end of the range, I guess. I bought some recently (Mouser #859-LTL2P3TBU2KS) that claimed 472nm but they're still a pretty deep blue. I also bought a few that were right around 500nm (Mouser #696-SSL-LX5093UEGC and others) but they all appear distinctly green to my eye (note, though, that I often disagree with my photographer girlfriend whose color acuity is probably much better.)

[Ian.M]

http://ledmuseum.candlepower.us/specx90.htm
Spectra for a few SiC LEDs from the same era.

[edavid]

I remembered the SiC blue LEDs as being more "magical" also, but when I actually dug them out and compared them to modern GaN blue LEDs, I found the visual difference was very subtle.

[Someone]

http://ledmuseum.candlepower.us/specx90.htm
Spectra for a few SiC LEDs from the same era.

Much broader emissions than contemporary LEDs:
http://ledmuseum.candlepower.us/specx107.htm

[Brumby]

(note, though, that I often disagree with my photographer girlfriend whose color acuity is probably much better.)

Yeah - you are probably going to lose that argument.

I've been working with an ongoing photographic job where reasonable colour accuracy is important.  (Being for internet viewing, this is somewhat of a futile exercise ... but I have to be as close as reasonably possible.)  In that exercise, I have come to appreciate colour variations to the point where I can often pick what corrections are necessary in one or two steps.

It's a sensitivity I never expected to develop - and it sometimes shows through at moments where ignorance of such things is the norm.   

[floobydust]

The first blue LEDs were SiC based and very low efficiency.

With modern GaN based LEDs the color depends on the exact type and can be somewhere in the 400-480 nm range.

I still remember the color of the very first blue laser pointer (around 1999) - a faint blue -violet at around 405 nm. Not very bright unless it hits something with a florescent dye.

I bought one Siemens SiC blue LED in 1985 I think it was, very expensive $60 at the time.
But super cool, it looked just great after decades red, green, yellow, orange LED's. 
Can't remember the forward voltage, I thought I ran it off two NiCd's, so it must be low Vf.

[Tomorokoshi]

(note, though, that I often disagree with my photographer girlfriend whose color acuity is probably much better.)

Yeah - you are probably going to lose that argument.

I've been working with an ongoing photographic job where reasonable colour accuracy is important.  (Being for internet viewing, this is somewhat of a futile exercise ... but I have to be as close as reasonably possible.)  In that exercise, I have come to appreciate colour variations to the point where I can often pick what corrections are necessary in one or two steps.

It's a sensitivity I never expected to develop - and it sometimes shows through at moments where ignorance of such things is the norm.   

When school pictures were done on film, the processing labs had (pretty much exclusively) ladies who would key in color correction factors like +2 red, -1 blue while viewing the negative.

[Cyberdragon]

Are there photos of these old LEDs lit? I've only been able to find ones of them off.

[frozenfrogz]

Maybe this? http://ledmuseum.candlepower.us/led/1990.htm

[rdl]

I have a couple of blue LEDs I bought from Digikey when they first became available commercially. I think they are from Cree and I think they are SiC type. I know they were expensive. While I don't remember the exact price, in single quantity they were over $4 each. The color is not too different from a more modern blue LED, but I would say definitely less green. The big difference is brightness. With 5 volts at 7 mA current they are barely visible in normal room lighting. Newer blue LEDs can be almost too bright to stare at with that much current.

[Cyberdragon]

What material are the turquoise LEDs? The ones above 480nm.

[floobydust]

There many "art" blue-led's, which are a blue LED + phosphor to make a lighter or greener blue, or purple.

Kingbright offers special colours in SMT. I have tried them all.
Otherwise, eBay or Ali to get "ice blue" or "pink" (which is purple really, purple LED's are totally UV).

LED colours are kinda boring now, not many high brightness varieties.

[Cyberdragon]

There many "art" blue-led's, which are a blue LED + phosphor to make a lighter or greener blue, or purple.

Kingbright offers special colours in SMT. I have tried them all.
Otherwise, eBay or Ali to get "ice blue" or "pink" (which is purple really, purple LED's are totally UV).

LED colours are kinda boring now, not many high brightness varieties.

As far as I know real turquoise LEDs have no phosphor, unless it's really small. They seem to range from 485nm to 505nm. They run on 3.2V so it's probably some form of doped gallium. Not sure if it's GaN or not though.

https://www.superbrightleds.com/moreinfo/through-hole/5mm-aqua-led-505nm-t1-34-through-hole-led-w-15-degree-viewing-angle/4044/

[floobydust]

I know those as "cyan" LEDs which are short-wavelength (505nm) binned GaN pure green or emerald green LEDs. I think they are rejects for traffic light applications which are binned around 520nm I think. They have no phosphor, and no so turquoise looking.

[jolshefsky]

As far as I know real turquoise LEDs have no phosphor, unless it's really small. They seem to range from 485nm to 505nm. They run on 3.2V so it's probably some form of doped gallium. Not sure if it's GaN or not though.

I bought some of those and they looked really green to me. Different from old green LEDs but definitely green (this is the color my girlfriend and I disagree on the most—maybe I'll show her and see if she thinks it looks turquoise.)

I think they are rejects for traffic light applications which are binned around 520nm I think.

Based on personal anecdotes, I agree: sometimes traffic lights look a little bluish to my eye, especially if they're next to an incandescent green that's more yellow.

Unrelated side note: I noticed at one point that I was driving under a lot of traffic signals that would turn yellow just as they reached the top of the windshield from my point of view. I thought it uncanny until I realized my peripheral vision was colorblind. I thought I was very special until I learned that basically everyone has colorblind peripheral vision and it's just our brains filling in the details all the time. I can notice the effect with red or green LEDs that blink rarely (e.g. smoke detectors) and when I see them in my peripheral vision they usually look gray but sometimes in color that's randomly unrelated to the actual color.

[Cyberdragon]

So basically cyan LEDs are any LEDs that fell somewhere between blue and green? I think the closest to turquoise you could get would be the 490-495nm ones (the 505nm does look green).

[David Hess]

I bought some of those and they looked really green to me. Different from old green LEDs but definitely green (this is the color my girlfriend and I disagree on the most—maybe I'll show her and see if she thinks it looks turquoise.)

There are two different phenotypes for the gene which encodes the green photoreceptor protein and they are sensitive to slightly different peak wavelengths so different people see a slightly different green.