LED grow lights... input please.

[khronik]

He's right. LEDs are way more efficient than incandescent bulbs, but they're not 100% efficient. Here's the chart:

Luminous efficacy - Wikipedia, the free encyclopedia

An HPS bulb has a luminous efficiency of 22%, while white LEDs have efficiencies from 3.8% to 10.2%, with only prototypes reaching up to 22%.
So, by using LED bulbs for growing, you are spending more money per bulb, and you are going to be spending more money on your electricity bill.

Sounds pretty stupid to me.

You're missing the one major advantage of LEDs here, that being that LED systems only output light on wavelengths that is ideal for plants to use. So even if a HPS system is 22% efficient, about 90% of that light is wasted because the plant can't use it, so really only 2.2% of that light is being used. An LED only puts out light one a single wavelengths, so with them, the numbers are reversed, and like 95% of the light being output is absorbed.

I found another site selling LED grow lights, this one looks a little more promising:

gro-tek.com

[Al B. Fuct]

LEDs produce a very narrow spectral output; a bandwidth of only about 25nm wide for the most part. LEDs might be able to produce the centre-frequency of an ideal flowering spectrum and a few nm each way, but very little outside that narrow bandwidth, contributing to a completely insufficient luminous output overall.

Have a look at the broad spectral output of the sun (which is what we're shooting for), the narrower and peakier HPS (though which is reasonably well matched to the plant's flowering requirements) and then these sample red, blue and yellow LED spectral curves. Because of the narrow bandwidth of current technology LEDs, if you were replicating Old Sol or even HPS, several LEDs with slightly differing centre frequencies in each main color band would have to be used- and each would have to produce high luminous output, around 8-10K lumens each, to be of any effectiveness, even compared to CFLs.

If all that was needed to grow dope was light of the correct colour, you should reasonably be able to make up an image in Photoshop which is the correct colour and point your monitor at the plants... and I think we all know that won't work.

LEDs are a great technology. One of these days, someone will work out how to get the luminous ouput level and breadth of spectral output where they're needed. However, until then, LEDs are toys which won't actually grow you any appreciable amount of weed.

Experiment all you like- but by the time you get done sinking several hundred bucks into LEDs, you'd have had an HPS and a decent exhaust/ventilation system bought and paid for.

[Al B. Fuct]

You're missing the one major advantage of LEDs here, that being that LED systems only output light on wavelengths that is ideal for plants to use.

Unfortunately, plants don't live on just a couple of narrow peaks in the spectrum.

Yes, cannabis prefers a blue emphasised spectrum for veg and a red-orange spectrum for flowering- but notice I said spectrum, not wavelength.

When we say wavelength, we're referring to a pronounced, narrow bump in the spectral curve around a central frequency. 'Spectrum' refers to a broad band of different wavelengths between an upper and a lower limit, with light of relatively equal amplitude (intensity) between the points.

Back to first principles. You can bet that the plant has evolved to grow in sunlight. Whatever lighting system you cook up, the closer it is to Old Sol, the more likely cannabis will like to grow under it. The sun is a broadbanded light (and other EM radiation) source. Current tech LEDs are not.

We know HPS is broad enough and produces enough luminous output in suitable ranges to flower cannabis effectively (and is no slouch vegging, either). Fluoros are broad enough but lack intensity. LEDs lack both breadth of spectra and output intensity.

It's gonna happen- someone will soon cook up LEDs bright and broad enough to be car headlights. When that happens, LED grow lights effective for cannabis might start to become a practical possibility.

[khronik]

Unfortunately, plants don't live on just a couple of narrow peaks in the spectrum.

They can. If they couldn't, LED grow lights would not work at all. Just because I don't live on pop-tarts, beef jerky, and vitamin pills, doesn't mean I couldn't.

Yes, cannabis prefers a blue emphasised spectrum for veg and a red-orange spectrum for flowering- but notice I said spectrum, not wavelength.

Yes, chlorophyll can absorb a spectrum of light, but i doubt it can tell the difference between different shades of red. It can tell the difference between blue and red, but that seems to be all. And LED systems are set to use the wavelengths that are most readily absorbed in the blue and red spectrums.

When we say wavelength, we're referring to a pronounced, narrow bump in the spectral curve around a central frequency. 'Spectrum' refers to a broad band of different wavelengths between an upper and a lower limit, with light of relatively equal amplitude (intensity) between the points.

Back to first principles. You can bet that the plant has evolved to grow in sunlight. Whatever lighting system you cook up, the closer it is to Old Sol, the more likely cannabis will like to grow under it. The sun is a broadbanded light (and other EM radiation) source. Current tech LEDs are not.

This last part I take issue with. Does any indoor grower try to emulate the sun with their lighting system? No, they try and emulate the chlorophyll absorbsion spectrum. Plants, and all other living things, have evolved to survive in the natural environment, but that doesn't mean that environment is ideal for them. I for one, prefer my house to the african savannah.

I attached a nice graph of an LED grow light spectrum superimposed on the chlorophyll absorbsion spectrum.

[Al B. Fuct]

OK, khronik, if LEDs work so well, you show me pix of an LED grow that's doing as well as an HPS op with all other conditions right.

They just don't make the luminous output needed- this is why LED grows look like they've been left in the dark, because in comparison to what the plant needs, that's ostensibly what it's getting.

You won't mind if I don't chug Jack Daniels by the case while I'm waiting for pix of those forearm sized colas, willya?

I'm all for new tech. LEDs will get there- but they are not there yet. Give it 5 years and we'll talk again.

Mind you, I'm not taking YOU to task- only LED grow lighting technology.

[Schrodingersdog]

I think that using LED lights would work fine for growing pot. You just have to cover the absorption spectra of the plant. The only way, as far as I know, that the plant absorbs light energy is through chlorophyll a and b.

Looking at the absorption spectra chart Chlorophyll absorption spectrum It looks like there are peaks around 425, 460, 640, and 660 nanometers. LEDs are specified by their dominant wavelength...easy enough to select some that would work for that (or at least close to it...they only produce them at certain frequencies after all). As long as these peaks are covered, the plant will gather energy, probably not as much as from a full spectrum light, but the light that strikes the plant will be more efficiently turned into energy for the plant to use. HID lights do the same thing, but there are a lot of wasted frequencies emitted that the plant uses very inefficiently. Check out this link if you want to learn more than you ever cared to about light emission spectra: The Science of Color, the Emission Spectra of the Elements and Some Lamp Engineering Applications

Ok...so what about the brightness of LED lights as compared to HID lighting? It is true that most LED lights are not very bright (~.5 lumens each), but certain types are produced (do a search for superflux LEDs on ebay) that put out ~40 lumens each. Thats pretty good when you consider that an array of 50 puts out about the same as a 100W flourescent bulb (~2000 lumens). Yes, thats not nearly as much as, say a 400W HPS (~58000 lumens), but if heat and power consumption are a concern, then LEDs may be a viable option.

I built two little LED arrays out of the superflux LEDs I talked about and currently have a few seedlings underneath them. They are at least not dying! I'll upload some pictures soon. My aim is not really to have a large yeild necessarily, but to reduce heat and power consumption...and to try something cool. They were pretty cheap to build by the way...all it consists of is 2x $10 packs of LEDs, some perfboard (~$10), some reflectors and a DC supply (from an old computer....dedicated DC power supplies are pretty expensive) I jacked from work. It did take a while to wrap the wires however. Who knows...I haven't done a lot of growing and chances are I'll kill my plants without seeing how well my LEDs work!

[oldsanclem]

This has gone on for a few years now and not one, bit of data.
Grams/watts/month. If It worked It would be used by NASA , they have more money than any pot grower. But they still use florsent pannel arrays.
O God I want it Fast, Cheap, and right now.
You want to know something funny a 30 year old ballast and new bulb, still is the proven way. Wow ballast older than most on this board O Well

[khronik]

To be fair, NASA is testing LED grow light technology.

The new digital ballasts are an improvement on magnetic ones though.

[Schrodingersdog]

Just an update on my seedlings...they're alive and well under the LEDs, but not having grown too much I can't say how they're doing in comparison to HID lighting. They've stretched a bit till now they're pretty much touching the lights, but there's not enough heat put off by them to harm the plants it seems.

Soon I'm gonna hook up a wattmeter to my power supply and see how much power I'm actually drawing, but ideally it should just be about 3.2(Volts per LED)X.02(amps)X50(#of LEDs) = 3.2 watts!

Since its a ghetto rigged power supply, some is being wasted though a dummy load to trick the computer power supply into thinking there's a motherboard there.....so we'll see how wasteful these suckers are....

[Opie Yutts]

So, the seedlings should have grown now. What do the adults look like? Please?