LEDs for lettuce and microgreens

I have been trying to grow leafy plants under fluorescent lights but still have stretching and limp leaves.

I have seen light integral numbers of 12-13 mol m2 d and Cornell has it listed as 15-17. I haven't figured out how to get this number for fluorescent or led lights.

Cornell says that 6 x 32watt fluorescent lights in a 2ft x 4ft area is adequate. They also say a 400watt hps will cover a 4ft x 4ft area. The operating cost of these at 14 hours is around $45 a month, ouch.

How would one figure out led coverage for a 4ft x 8ft table? I'm trying to compromise between operating cost and upfront cost of cobs, drivers and heatsinks.

These are in a temperature controlled room with a/c so additional heat load effects summer operating cost, I'm thinking leds should lower my air conditioning cost.

I looked at them based on lumens to the lumens of the fluorescents. It looks like the Vero 29 would only save me 30 odd watts. The vero 10 would save me 60 watts but would have to purchase 14 of them. So I'm hoping there's a better way to figure out how many I need for max efficiency to beat out using flourescents

The Vero 10 should work great. This one is rated at 126 lumens/watt at 350ma.
http://www.digikey.com/product-detail/en/BXRC-40E1000-B-03/976-1173-ND/3913159
14 of them should make about 16500 lumens. 4000k is an ideal temp for what you're trying to do.
At $4.64 each 14 would run $64.96 + shipping from Digikey. For a compatible driver I recommend the Meanwell HLG-120H-C350A. It can power up all 14 emitters on a single string and is dimmable which comes in handy.

Thanks. What heatsink would you use? I would want to spread them out evenly for coverage so thinking individual. The arctic ones used for the cree seem like overkill but don't know if there would be a cheaper solution. Thanks

That only saves me 62 watts over the fluorescent for lighting. Not sure how to tell what it saves me in heat load.

Was hoping the led wouldn't require as many lumens to get the job done :/

ROF42 said:

Thanks. What heatsink would you use? I would want to spread them out evenly for coverage so thinking individual. The arctic ones used for the cree seem like overkill but don't know if there would be a cheaper solution. Thanks

That only saves me 62 watts over the fluorescent for lighting. Not sure how to tell what it saves me in heat load.

Was hoping the led wouldn't require as many lumens to get the job done :/

Click to expand...

With the Vero 10 at 350ma you can passively cool them. I have 3 mounted on a 4" x 12" heatsink ~3/16" base and 8 3/4" tall fins.in my clone chamber with no fans. The heat sink gets warm but not hot. You could use a 3" piece is this http://www.heatsinkusa.com/4-230-wide-extruded-aluminum-heatsink/ for each emitter without needing fans. If you want to run them on separate sinks you want separate drivers. I would look on ebay for a driver capable of powering 10 1 watt leds for each emitter. Sometimes they can be found for a few dollars each but they do ten to run less than 350ma. More like 280-300ma. And, they won't be dimmable.

Led does tend to make more PAR lumens per watt as well as much better lumen maintenance over time so it's pretty likely you will not only have a lower power bill but better results and less heat too.

Here's a cheap driver that will power one Vero 10 at 300ma.
http://www.ebay.com/itm/6-10-1W-LED-External-Driver-Power-Supply-85-260V-Short-Circuit-Protection-/271316111004?pt=LH_DefaultDomain_0&hash=item3f2bb34e9c

$2.09 each

How about this driver?

http://m.ebay.com/itm/371049866835

Just saw yours and it's cheaper. Thanks

14 x Vero 10 $64.96
14 x Drivers $29.26
14 x Heat Sink $60.48
----------------------
$154.70 to cover 2ft x 4ft

$618.80 for one table, ouch.


14 x 9.345w = 130.83 watts

6 x 32w fluorescent = 192 watts

Saving 61.72 watts * 14 hrs a day * 30 days a month / 1000 to kilowatts * $0.14 per kilowatt

Saving $3.60 a month
Pays off in 172 months.

I was hoping I could find numbers that said I needed less led power. Seems like they just aren't far enough along to be worthwhile.

I do however plan to use these in a terrarium. I'll see how succulents do under one or two, and then see how many a micro rose bush needs

ROF42 said:

14 x Vero 10 $64.96
14 x Drivers $29.26
14 x Heat Sink $60.48
----------------------
$154.70 to cover 2ft x 4ft

$618.80 for one table, ouch.


14 x 9.345w = 130.83 watts

6 x 32w fluorescent = 192 watts

Saving 61.72 watts * 14 hrs a day * 30 days a month / 1000 to kilowatts * $0.14 per kilowatt

Saving $3.60 a month
Pays off in 172 months.

I was hoping I could find numbers that said I needed less led power. Seems like they just aren't far enough along to be worthwhile.

I do however plan to use these in a terrarium. I'll see how succulents do under one or two, and then see how many a micro rose bush needs

Click to expand...

I might mention that in 172 months fluorescent bulbs will need to be changed like 12 times. That should be part of the calculation. Also, 4 32w fluorescent bulbs output about 12800 lumens when new. That works out to 100 lumens/watt. The Vero 10 4000k outputs 126 lumens/watt. 26% greater efficiency with much longer lifespan are among the main reasons led is a good choice over fluorescent lighting.

All true, but part of that only matters if they're still in use 14 years from now. Seems like forever away but sure hope my business is going strong at that point.

I was hoping one per square foot would work.

I see Aquarius has the vero 10 a 350ma listed as 3.07 under par@50 (watts). I've been trying to find a par number in watts for the fluorescent bulbs to compare but no luck yet

PPFD x hours x 0.0036 = DLI

Using 24 hour photoperiod, you would need 200 micromoles for PPFD to achieve 17 mol DLI. 130 lumens/watt with vero or cree leds converts to roughly 1.8 to 1.9 micromoles/watt/s. 1 square meter is roughly 11 sqft, so 18 micromoles/s per squarefoot for flux. So roughly, 10 watts per sqft with 130lm/w leds should do the trick. Hope that helps.

You might find this useful too:


http://spectra.1023world.net/

anomuumi said:

PPFD x hours x 0.0036 = DLI

Using 24 hour photoperiod, you would need 200 micromoles for PPFD to achieve 17 mol DLI. 130 lumens/watt with vero or cree leds converts to roughly 1.8 to 1.9 micromoles/watt/s. 1 square meter is roughly 11 sqft, so 18 micromoles/s per squarefoot for flux. So roughly, 10 watts per sqft with 130lm/w leds should do the trick. Hope that helps.

You might find this useful too:


http://spectra.1023world.net/

Click to expand...

Are you saying leave the lights on 24/7?

I'm going to attempt to work your post backwards and see if I can figure out the formula. It looks like lettuce is usually grown with 14 to 16 hours of light. I'm not sure if this is bc lettuce needs a dark cycle, or that's where they tend to get the daily light integral from fluorescent lights.

Also will see what I can come to lowering DLI to the 12-13 range.

Trying to make it economically viable. I'm in the south where it's just too hot in the summer and cold enough to slow growth in winter on outdoor crops.

Thanks

The cheapest HeatsinkUSA profile/surface area is the 2.08". It would be perfect for your project because you do not need strong intensity and it will space the Veros out quite a bit, while still giving you enough surface area for passive cooling. That would help reduce your up front cost.

Typical fluoro is only 20% efficient and the best value fluoro might be lucky to get 25%. The Vero10 5000K at 350mA (9.2W ea) is 40.4% efficient. As you probably know, fluoro bulbs experience significant lumen depreciation over time but the Vero setup would experience virtually no lumen depreciation over time, it may even appreciate. So with 20% efficiency, we also have to factor in reflector losses, which are very substantial for fluoro. It is just not an ideal way to get photons into the canopy. The photons need to bounce too many times and the bulb itself blocks its own light. The Veros could be run close enough (soft enough) that they would not need a reflector. If you choose to run them harder and raise them high enough from the canopy that the DO need a reflector, it will be much more effective at directing the photons to the canopy than the fluoro reflector.

So long story short, you can use much less emittance to do the same job (about 3X in this case). The Veros will give you twice as much light and more of that light will get into the canopy. Combining that with excellent lumen maintenance, and the cost may seem more worthwhile.

For reference, in my vegging space I use about 8 dissipation Watts per ft² to get vigorous vegetative growth with a lamp efficiency of 45-50%.

SupraSPL said:

The cheapest HeatsinkUSA profile/surface area is the 2.08". It would be perfect for your project because you do not need strong intensity and it will space the Veros out quite a bit, while still giving you enough surface area for passive cooling. That would help reduce your up front cost.

Typical fluoro is only 20% efficient and the very best fluoro might be lucky to get 25%. The Vero10 5000K at 350mA (9.2W ea) is 40.4% efficient. As you probably know, fluoro bulbs experience significant lumen depreciation over time but the Vero setup would experience virtually no lumen depreciation over time, it may even appreciate. So with 20% efficiency, we also have to factor in reflector losses, which are very substantial for fluoro. It is just not an ideal way to get photons into the canopy. The photons need to bounce too many times and the bulb itself blocks its own light. The Veros could be run close enough (soft enough) that they would not need a reflector. If you choose to run them harder and raise them high enough from the canopy that the DO need a reflector, it will be much more effective at directing the photons to the canopy than the fluoro reflector.

So long story short, you can use much less emittance to do the same job (about 3X in this case). The Veros will give you twice as much light and more of that light will get into the canopy. Combining that with excellent lumen maintenance, and the cost may seem more worthwhile.

Click to expand...

When you say 3x less, does this mean only 1/3 the lumens or 1/3 the watts one would need for fluorescents?
Thanks

And would you just do 4ft sections of that heatsink?

Supra, how many hours do you leave the veg lights on?

Yes I estimate you could use 1/3 the dissipation wattage. That is based on my assumption that the fluoro suffers 33% reflector loss and the Veros lose 5% to scattering.

If you have large space to cover, the long bars could be convenient, less wires and less heatsinks to hang. But it will limit your height adjustabiltiy so that is up to personal preference.

Also worth mentioning, I have not tested the Vero 10 on this heatsink profile yet. I have some on order for a vegging lamp design so if you are willing to wait for a week or so I can report back with exactly how they perform in terms of heat and light spread. My plan is to run them at 300mA (8W), spaced 12" apart. Using this driver

Thanks. I'm still in the planning stage. I was rushed through my last attempt, not allowed to fully research and plan due to impatient business partner. It has turned out to be pretty much a disaster so this time I'm looking to fully research my needs and costs first.

I really like the idea of vero 10s ran passively as a replacement for fluorescent tubes. Planning on making a few myself ran at 300mA.

I know nothing from growing lettuce, but longer photoperiod would be an economical way to achieve higher DLI with lower equipment costs. Following sources might provide some insight too:

http://europepmc.org/abstract/MED/11539771

http://hortsci.ashspublications.org/content/33/6/988.short

ROF42 said:

I have been trying to grow leafy plants under fluorescent lights but still have stretching and limp leaves.

I have seen light integral numbers of 12-13 mol m2 d and Cornell has it listed as 15-17. I haven't figured out how to get this number for fluorescent or led lights.

Cornell says that 6 x 32watt fluorescent lights in a 2ft x 4ft area is adequate. They also say a 400watt hps will cover a 4ft x 4ft area. The operating cost of these at 14 hours is around $45 a month, ouch.

How would one figure out led coverage for a 4ft x 8ft table? I'm trying to compromise between operating cost and upfront cost of cobs, drivers and heatsinks.

These are in a temperature controlled room with a/c so additional heat load effects summer operating cost, I'm thinking leds should lower my air conditioning cost.

I looked at them based on lumens to the lumens of the fluorescents. It looks like the Vero 29 would only save me 30 odd watts. The vero 10 would save me 60 watts but would have to purchase 14 of them. So I'm hoping there's a better way to figure out how many I need for max efficiency to beat out using flourescents

Click to expand...

Do you have any t8 fixtures laying around??? can look at cree T8 retrofit led bulbs(21w/2100lm/4ft) as another option for growing micro greens. Their around $30-34 a piece though until HD finally carries them?!, still waiting

I have the clips for them on 1" x2" over my current indoor attempt with ballasts.

I think I'm set up with 12 lights right now.

How would this compare to the v10? I'll have to write down the numbers