Do super-red LEDs emit 'pleasing' light?

As far as I remember, red leds have always been emitting at 660nm, with 1.6V drop.

Meindert

Hello Meindert,

Most of the ones I have used were 635nm. I don't know why they call

660nm "super red". I know that 700nm used to be called "dark red".

Regards, Joerg

For illumination, probably not. For annunciators, absolutely! :-)

For illumination, I'd go for green.

Good Luck! Rich

2V ? @ what current ?

The ones I use are typically around 1.6V.

Can you point to a data sheet for the 2V ones ?

Graham

Hello Rich,

I would have gone for green as well but with 2 AA cells that just isn't going to work. Usually I like to reserve the color red for hard alert functions. Yellow for warnings and red for when something has gone wrong.

Regards, Joerg

660nm GaAlAs is super (bright) red, not super-red in color.

635nm/640nm GaAsP/GaP is called High efficiency and is orange-y color. There's also 655nm GaAsP and 700nm GaP. The dimmest is usually the

655nm GaAsP or the 700nm GaP/GaAs. Typical Vf varies about 1.7 to 2.0 at 20mA (call it 1.85V +/- 150mV typical), depending on type and die size.

Best regards, Spehro Pefhany

--
"it\'s the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

Hello Graham,

That's at full brightness. I am going to limit it to whatever is acceptable since it will be the lion's share in battery consumption.

Regards, Joerg

Hello Spehro,

Thanks for explaining. For some reason Lumex calls their 660nm versions "super red":

What I don't like is the fact that they distribute uncontrolled documents. Says so on top of it. In medical electronics we'd be in hot water if we ever did that. The Federales would be all over us.

Regards, Joerg

It's industry-standard terminology, at least in Asia.

It's just an LED, lighten up. ;-)

Best regards, Spehro Pefhany

--
"it\'s the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

I much prefer 635nm (helium-neon laser color) to 660nm. By my eyes

635nm is a pleasing "red." 605nm is "orange" to me, with 595 appearing as "dirty amber" and somewhere around 585-590nm appearing "amber."

Don't forget -- the eye is roughly 3-4x more sensitive to 635nm than to 660nm, at least according to this chart I'm looking at, so you might be able to use a 635nm part and save current, run the LED at lower Vf, use a cheaper, lower-output-spec LED without losing apparent brightness (yes Joerg, we remember your fondness for cents-i-ble savings :-), or any combination of these.

Best, James Arthur

Hello Spehro,

Yeah, I know. It's just that after working in med for 20 some years this kind of stuff leaves a stale aftertaste.

Regards, Joerg

Hello James,

Thanks, I didn't know this. 3-4 is a lot. This time it's less about cents and more about yielding long battery life. I could also flash them but that can cause fatigue and strain.

Regards, Joerg

Human eyeballs are indeed dramatically more sensitive to 635nm photons than they are to 660nm photons. However, this eye sensitivity difference is already taken into account when comparing different LED "mcd" or millicandela ratings. A candela is a lumen per steradian. A lumen is a measure of total light output compensated for the relative efficiency of the human eyeball over the entire visible spectrum. A steradian is a solid angle measurement.

So... A 1 lumen 635nm light source should appear to a human to be of approximately equal total light output as a 1 lumen 660nm light source. Similarly a 40mcd LED with a viewing angle of say 110 degrees at 635nm would also put out approximately the same human perceived total light output as a

40mcd 110 degree 660nm LED.

As it so happens the Lumex LEDs you linked to have specifications of 40mcd @

60 degree viewing angle @ 635nm, compared to 80mcd @ 110 degree viewing angle @ 660nm. Based on these numbers, the 660nm 80mcd 110 deg. LED is dramatically more electrically efficient and should also provide noticeably superior human perceived output as well.

Neither of these LEDs however produce anywhere near the best human perceived efficiency that technology currently allows. LEDs such as these will likely provide dramatically superior human perceived efficiency:

All of the available LED colors with the exception of many white and some amber, yellow, yellow-green LEDs are very pretty in my mind.

Something is puzzling me about these voltage-wavelength ratings.

A 660 nm photon has 1.88 eV of energy, so it should require at least

1.88 V to produce them in an LED. 1.7 V should only produce 729 nm (or longer) wavelengths. What am I missing here?

Regards,

Mark

p.s. for those who are not up on their quantum mechanics:

(photon energy in eV) = 1240 / (photon wavelength in nm)

Joerg,

Since you seem to be looking for better efficiency, have you looked at other vendors? Osram has a line of LEDs that pack a lot of punch for minimal current (100's of mcd for 20-40mA). They've also got a pretty wide range of colors available (~9, I think), and side-emitting packages too. I checked one red spec'd at 1.8 Vmin, so not sure if they'll meet your voltage budget.

DigiKey stocks a small portion of their line, and IIRC the prices are in the 35-cent range for red/orange/yellow/amber varieties.

- download the "short-form catalog" to get a good overview of their products.

Richard

LEDs are neither monochromatic nor "mono-voltaic". They produce *about* 660 nm at *about* 1.8V - and note that the current through them plays a part too. Vf needs to be specified at a specific current. It would seem unreasonable to expect a single frequency of light from one from say 2 mA to say 30 mA.

Ed

You could PWM the LED driver with a slowly changing sine wave (sort of like an ocean wave timing) which is very pleasing to the eye.

Jim

google for "LED switching supply", there you will find a led drive IC to eficiently drive a led or led chain ,no serial resistor other then a current sense resistor. Battery voltage can vary over a large range.

Thanks Fritz. I should've known that anything named after a candle would take the eye's response into account.

Yes, the flux/angle stuff I did know ;-) It would be nice if the makers would spec an overall luminous flux, making luminous efficacy easier to figure.

Not me...I didn't give any links.

Toshiba and Cotco LEDs were among the most efficient I found last time I looked (a year ago).

Lumileds' Luxeons are about the best-colored white LEDs I've seen thus far, but the standard parts don't render red all that well. Not horrible, but rather like an exaggerated version of the 'cool' fluorescent lights. I've not seen the 'warm' white versions--maybe they're better.

Some of the Chinese eBay-type 5mm white LEDs are very pleasing in color, but have rapid degradation of output intensity. Nichia's NSPW500C is rugged, but bluish still.

Best regards, James Arthur