I purchased 30 pcs 10w red and blue LED chips. After this unlucky auction I found out not only that the BuyItNow price was lower, but that the LEDs weren’t what I expected; they were ordinary white LEDs with a color gel filter on them - no green energy for my flowers:(.
Now I ponder how I could change the color (or invisible spectrum) of light. Even though accelerating my LEDs forward and back at my plants to red shift and blue shift their light would not be practical - any suggestions are welcome.
Can someone calculate how much energy I waste with the gel filter? My LEDs take 10V and 1050mA. Luminous intensity of a white LED is about 900 lm, but how much I will get?
And sorry, but you can’t change that. LEDs work on a quantum level emission/absorption interaction. The only way to alter that is to change the laws of physics for the entire universe.
It would be more productive to ask how to change lead into gold so you could buy some real grow-lights.
Surely that is ideal for a grow light, then? Leaves look green because they absorb less green light than any other wavelengths. In other words, they are absorbing the red and blue parts of the spectrum.
Well, it’s usually a relatively sharp blue and a broad “red”. There’s still going to be an overlap with the absorption of the plant’s leaves, so the suggestion to just let it all fall on the plant is clearly best.
I observe, by the way, that although our eyes peak response is in the middle of the green, the reason plant leaves are green is because they reflect green light (as colophon observes). Here’s the absorption spectrum of chlorophyll:
You CAN alter the wavelength of light – but you’ll end up either absorbing the light into something and re-emitting at a longer wavelength (losing part of the energy in the process, as longer wavelength photons have lower energy – this is whjat that phosphor is doing to the blue LED light in that “white” LED), or else you have to use some nonlinear optical process that requires pretty high input per unit area to begin with (the most familiar example being those green laser pointers, which actually use an infrared laser diode and a frequency doubling crystal). In either case, you’re not going to get efficient transforming of LED light to a different color
It’s not Blue and Red – it’s a relatively narrow blue LED (although the spectrum given looks rather broad) and a broad phosphor that covers a lot of wavelengths. By itself, I believe the phosphor looks kind of yellow. The combination of the two gives you a sort of bluish white. I think they use them in a lot of those auto headlights that are an annoyingly bluish color.
The thing is, there are lots of phosphors you can use. Companies like Osram/Sylvania have vast “libraries” of such phosphors, and can give you all sorts of color balance. But the generic “white” LED available today from electronics suppliers still tends to be this awful soul-less bluish white. I’ve seen Osram’s “warm” phosphors that give you something closer to an incandescent lamp, but they’re not generally out on the market. I think the best thing companies like Osram could do would be to make such LEDs more widely available – it would do a lot to alter the public perception of what LEDs can look like.
Well, not quite. As I state above, green laser pointers use a frequency doubling crystal to shift infrared light to green by doubling, since 2 X 550 nm is 1100 nm. You can also do frequency tripling or quadrupling. But it’s harder to do a fractional shift of the sort that would be needed to shift red to green.
One reason it’s hard to do this way, and requires very high power concentration, is that you need to combine two photons of infrared to make green, since each green photon has twice the energy of the infrared photon. You need to be sure you’ll have the two photons in the same place at the same time. Until lasers were around, such nonlinear frequency doubling and frequency combining was such a small effect that it wasn’t observed.
Going the other way – green to red – is easier. All you have to do is have your green light absorbed by a phosphor, which spits out the red photon, and turns the rest of the energy mismatch into vibrational energy (i.e. – heat)
No. It’s easy to change the frequency of light emitted by a LED, all you have to do is change the temperature: the frequency of the light is determined by the energy gap of the semiconductor, which is the energy that is released upon recombination (the electrons ‘dropping’ from the conduction to the valence band). This energy gap changes as a function of temperature, because the temperature change influences the lattice parameters of the semiconductor (i.e. it makes the atoms come closer together/move farther apart on average). Basically, if you heat it up, the band gap shrinks, so does the energy released in recombination, and thus, the released photons will have higher wavelength/lower frequency.
So, no changing of the laws of physics or other alchemy needed. However, the effect is probably to small for the OP’s needs, and I’d WAG it’s likely you kill the LED before you get any significant color change.
So you can’t make any noticeable change, and any even an attempt at making an *unnoticeable *change will destroy your LED. But it’s possible. :dubious:
Yep, and the operation was a success, but the patient died. Classic doublespeak.
You can certainly make a noticeable change, even if not a very big one. According to this source, wavelength change varies between 0.03 and 0.13 nm/°C, while the just noticeable difference for light wavelengths is between 1 nm at the peak, and 10 nm at the edges of the visible spectrum. So a 10°C difference can produce a noticeable change in color under optimal conditions.
Any kind. We start all our plants under lights before moving them to the greenhouse or the garden. The power bill is not insignificant, but I haven’t done a CBA on switching sources yet.
No, you can’t change the frequency or wavelength of the light produced by an LED by altering the drive circuit.
Some LED applications, notably IR remote controls as used on televisions and other entertainment devices do apply a RF or ultrasonic modulation to the IR to distinguish it from background. The net output is then IR with side bands that extend +/- a few hundred KHz from the LED’s intrinsic spectrum. Large LEDs have such long carrier lifetimes that you can’t possibly modulate the drive at a high enough rate to visibly alter the spectrum.
Avoiding confusion like this is why I find it helpful to speak only of wavelength when talking about light, even though as a EE I tend to mostly think in terms of frequency. Mentally I make the distinction that EM waves in a dielectric or vacuum have wavelength, and signals in wires have frequency, which kind of works until I have to deal with antennas and transmission lines.
