Characterizing white LEDs for general illumination applications .PDF document

Where they of good quality? Osram claims 30,000 hours from their white LEDs. They do last longer, but the white component only lasts 30k hours.

gives

asserted

left

illumination.

white LEDs. They do

Yeah, that's the problem. The phosphor dims, just like the phosphor in a CRT when you have a pattern constantly on the screen in the same place. The bright spots get burned in, so you can read it even when you turn the monitor off. The same thing happens to the phosphor in the LED, it dims, and then there's little blue and almost no white.

THe LEDs I used were from Hong Kong, and no, they were cheap, so gauging by the price they were low quality.

ok, if you have them running on say a 50mhz signal pulse generated then they will last longer. I recon make sure you only apply enough energy to power them up not the max rated tollerance.

| Yeah, that's the problem. The phosphor dims, just like the phosphor in | a CRT when you have a pattern constantly on the screen in the same | place. The bright spots get burned in, so you can read it even when you | turn the monitor off. The same thing happens to the phosphor in the | LED, it dims, and then there's little blue and almost no white. | | THe LEDs I used were from Hong Kong, and no, they were cheap, so gauging | by the price they were low quality.

There are better phosphors these days that don't have that burn in problem. I guess cheaper LEDs don't use those.

However, I don't like the spectra I get from a CRT screen for illumination purposes. If the same spectra comes from while LED's I won't like them, either. There are LEDs that produce single discrete colors from all over the spectrum. Maybe the right phosphor mix can be made to do that for one single LED, too. A CRT would not use such a mix because it needs to have discrete colors. But a white LED in theory could, and could potentially emulate an incandescent light source very well. A pair of them could be used to balance the color temperature, too. Otherwise I guess I'll have to make a box with lots of different LEDs all over the spectrum, and set their levels to get the proper broad spectrum of light output as white.

Is this true? Do white LED's actually use phosphor to generate the white light? I assumed it was doping in the silicon junction that produced specific wavelengths of light just like red or green LED's. If phosphor generates the white light, what wavelength excites the phosphor? It can't be infrared (can it??). Do they make LED's that output in the UV region, and then put phosphor in front of the junction?

Thanks for the info.

Most "white" LEDs actually emit blue light from the junction. The red and green is generated from the blue light by phosphor in the plastic. Red + green + blue makes white in additive colors (as on a CRT screen). It's also possible to generate light that appears white by using red, green and blue LED chips, but that is a lot more expensive.

Best regards, Spehro Pefhany

--- The white LED's I have on hand have a water-clear plastic T1-3/4 housing, and when you look through it into the reflector cup (when it's not illluminated) you see a light yellow non-reflective coating on it, which glows bright white when the LED is on. The chip itself doesn't seem to be emitting blue, so I believe that what's happening is that it's generating UV which is being used to pump the blend of phosphors in the cup into emitting the three primary colors which we then perceive as white light.

-- John Fields

Yup... not unlike a standard fluorescent tube.

"White" isn't a color, but a combination thereof (as you realize), so you'd need several different dopings or junction chemistries in order to generate the necessary wavelengths. As far as I know, nobody's managed to do this sort of multiple implanting within a single junction, or on multiple closely-located junctions on a single die.

There are tricolor LED assemblies available, which have separate red/green/blue LED chips mounted on a single substrate, usually feeding into a single lens assembly. They're more complicated to drive, due to the fact that the different wavelengths require different bandgaps, which require different forward voltages - you can't simply wire the three chips in parallel, or the red one (which has the lowest forward voltage) will hog all of the current. They're also expensive.

Fun, though, since you can generate a wide gamut of colors by driving the three chips to different brightness levels (adjusting the current, or using PWM). You'll sometimes find these used in small consumer-electronics components, as indicators - the Yaesu VX-7 handheld amateur radio has a multicolor indicator which I believe is powered by one of these.

Last I heard, they were using a blue LED, and a phosphor which absorbs some of the blue and then emits yellow and a smaller amount of red.

Yes, you can get ultraviolet LEDs these days.

white

Yeah, it's not uncommon (sic) to see this stumbling block to easy reading nowadays. I would say "Yup... like a standard fluorescent tube."

I'm not trying to pick on Dave, it's a journalistic trend that seems to be "not uncommon" nowadays. I'm ashamed to see, above, that I've fallen into its trap..

produced

can't be

and

AE6EO

Yes it is phosphor. most "white" LEDs are actually Blue LEDs with phosphor. They are trying to get UV LEDs working well enough to use for white LEDs since the result is better light, but I don't think they have solved all the problems yet. TTYL

be

wrote

be

I doubt it. UV LEDs, while they exist, aren't yet "good enough" for widespread use in white LEDs. The LED you have is almost certainly blue. TTYL

The spectral curves usually show a big, sharp blue peak and a broader yellow. If it's not too diffused, you can often shine the spot on a piece of paper and see regions of blue and yellow, because the phosphor geometry is generally sloppy.

John

But you can also get white LEDs by mixing colours, OSRAM makes a chip LED that emits "white" and can mix and achieve millions of colours. The white component however is made up of primary colours, but it looks really good!

-- Myron Samila Toronto, ON Canada Samila Racing http://204.101.251.229/myronx19