LED Lab 2009

[Weezard]

There are many "ways"

There are many "ways"

ok dat link was no use.

Weezard,

show me the way. How do i set one of these up im really intrested.

15W. Ledengin LEDs are probably the easiest to wrangle.
I buy mine from Mouser electronics on-line.

Then get a coil of soft copper pipe. the kind they use for water lines to the street.
It is super soft and can be hand bent to any shape,
Squeeze the pipe in a smooth jawed vise almost flat. Leave the ends round so you can connect them to a water supply. You can water cool the leds with a pump and res. to save on water.
If you use the hydro res for water you will need an in-line "therminator" to cool the water.

Then sand one side of the pipe flat and smooth and buff it till it shines.
Use Arctic Silver heatsink glue and attach a 15 W. 660 nm, Led every 6 inches along the pipe. Use a Royal Blue 15W. led every third space for a 4:1 ratio.
8 reds and 2 Blues will give you close to 150W. of light.
I'd arrange them; RRBRRRRBRR
The leds are best driven by a constant current supply but a tightly regulated 12V supply will do in a pinch at 8 amps for 8 red leds in parallel.
The Blue leds will want very close to 15 V. and 2 amps. for the 2 leds.
This will outgrow a UFO or a Procyon for half the cost.
I think you can find the Reds for about $22 ea and the blues were $36 ea. last I looked. About $250 for leds and another $100 for copper and power supplies.
And you are good for 10 year of no lamp replacement and low power bills

That's one way.

Weezard

 

[Weezard]

Agreed

Agreed

cool, I'm subscribing to this thread. I've been planning on building my own LED system for some time now. Seems like key to a good LED grow is scrog/lst. I'm designing a 1.5 sf box build just for that.

Leds lack the penetration to grow tall bushes.

But with lst/scrog they really shine.
I'd like to see your design.
You show me yours.
I'll show you mine.

Aloha,
Weezard

 

[CabGrower]

This thread is getting good! Those links for the LEDs kick ass, they seem like a decent way to go and I could totally build that.

Does anyone know if we could use a computer power supply to drive them? They would need a resistor in series to regulate the current.

 

[knna]

I dont know if a disscusion about this topic is going to go anywhere: it has been tried some times before at this board. Im not going to post again sources, graph, etc.

If other experienced LEd growers chime in, ill see what i do. If there is mutual sharing of info...im really tired of posting about my research without any feedback, while seeing people using it to sell bad designed LED lamps without doing any own research.

So ill start giving just some personal opinions about the questions on your first post, ballast.

To begin with, why do so many LED grows reportedly fall short of HPS or CFL? Several possible reasons come to mind:

* misunderstanding of lumens, PAR, radiant flux, mcd, etc. how much usable light are plants actually getting? marketing hype may vastly overstate it.

* photosynthetic action spectrum and LED wavelengths - what quality of light are plants getting? it seems that more experimentation is needed in this area for cannabis, including UV and IR effects.

* less efficient older LED's - comparing watts alone hardly tells actual light output.

* cost - grower may tend to use less light than if cost were not a factor.

I cant agree more about this as the cause of so many failed LEd grows. Resuming the three points, people is using way less light than required.

There is little general knowledge about plants and light on all MJ forum at the net. Marketing hype and oversimplification, coupled with lots of myths. Ive been fighting against them for some years, but ive gave up. 99% of growers arnt interested on facts about light and plants. They prefer straigh and simple answers for a very complex field that almost never have such answers.

Plants use photons, not watts or lm. Number of photons (abreviated, uE) is by far the main parameter to consider for plant's growing.

Have you seen any grow reporting how many uE is using? Or light densities (uE/m2)? There is no way we can increase our knowledge about how much uE cannabis needs and what light qualities it prefers if we cant compare results based on the only unit meangliful for that task (uE).

So almost all you can find at the Mj forums about light and plants is lounge's talk. If a LED lamp designed on that works decently, it is going to be mainly luck, with little knowledge of why it works of not, so there are little possibilities of improving it further.

About LEDs to use, blue light efficacy for grow is vastly overstated. But the fact is it induces very low photosynthetic response, compared to red. Blue is required for health grow, but using more of what is strickly required for cannabis is just a waste of electricity.

Plants perform better when receiving light along all the PAR range. Cool White LEDs (CW) are good at it, giving enough blue,green and yellow. Simple CW plus red leds combinations are giving the best results, in terms of dry yield per uE. If wanting to reduce green and yellow on the overall spectrum, just adding some blues to reach the minimun blue content and reducing some CW are able to get strong red with a decent blue peak and some of the rest of the wavelenghts.

Nobody knows the spectrum distribution at what cannabis perform better. Ive found very little feedback of other LED growers about this esential topic on the last years, so i doubt it is going to change in the future. Too much economic interests, thought. Actually we dont know more about this than 3 years ago (apart of the obvious fact that red+blue mixed dont perform well, specially at the bloom stage).

When selecting LEDs, the most important is their energy efficiency (fraction of the burned watts that they emit as light), not peak wavelenght. 660nm leds use old AlGaAsP technology, that reach 22% efficiency on best cases (at 350mA). While AlInGap leds emiting on lower wl reach efficiencies of 30% (35% more light per watt burned) and increasing (660nm efficiency is stopped long ago). As using 640nm vs 660nm have a photosinthetic efficacy difference below 10% (actually, close to 5%), it seems a bad idea using 660nm.

Aditionally, the only high power 660nm leds i know are the LEDEngin and the Edixeon Federal (that is avalaible on 740nm, far red, too), that are comparatively expensive against Luxeon, Cree, Osram or SeoulSemiconductor leds.

On NorthAmerica, Luxeons are easily avalaible, but unfortunatelly currently their efficiencies arnt good (compared to others), especially on blue and white leds. Red leds top bins of all those manufactures have similar efficiencies.

 

[Supacalabala]

What do you need to get these 1w+ to work besides electricity resistors and a heatsink? I've wired tons of 5m leds but that was cake.

 

[Weezard]

Good to see ya KNNA!

Good to see ya KNNA!

Now you guys will get a schooling in LED and LED theory.
Knna knows his stuff!
His 2 cents is worth $8.95 and a bag of chips!
He is often way over my head,
but, I'm just a simple under-educated tiny country lizard.

You folks will probably understand the micro einsteins,
PAR radiation, flux capacitors and all.

If ya do, get your spectral advice from KNNA, then post yer choices for LEDs.
I'll try 'em and see if things get even mo' betta.
I'll be happy to help design constant current sources an heat sinking for anyone interested in building a grow light
with whatever spectra and efficiency you choose.
I will say, for all my lack of sophistication, I do get amazing results,
though I'm sure I waste a Watt here and there.
And I'm more willing than able to learn to do it better.


Now, If we can just get Physicsnole and Salmayo up in here
we can get back to work designing the "perfect" grow light.
Why? Because what's for sale commercially ranges from crap to very expensive "nice trys".
We're kinda stuck with rolling our own,ya?

Aloha,
Weeze

 

[knna]

Hello, Weezard

your experience building LEd arrays is what is going to be very useful for people wanting to start growing with LEDs. For my experience, building LED drivers is the weakest point currently on the proccess of making a DIY LED array.

People must be aware that LEDs are current driven devices, not voltage driven as most electronics components. Although it is possible to run a system using a fixed voltage Power supply (PS) and resistor or other ways of controlling by voltage, they are always very inefficients (losing on drivers any electrical saving gained on lighting). And as the current of the LED increases, it becomes exponentially more difficult.

I just would give an advice to people wanting to build DIY LED arrays: start thinking on how are you are going to drive it. Constant current drivers are the way to go. If a grower dont feel able to do it, must think on buying a prebuild one, and this add costs fast.

Electrical efficiency of drivers is not a problem for a experimental setup, but when wanting to use LEDs as alternative to other types of lighting, it becomes esential. Ballasts losses may account for a significant fraction of the used power.

Peace,
knna

 

[Weezard]

Spot-on!

Spot-on!

"Although it is possible to run a system using a fixed voltage Power supply (PS) and resistor or other ways of controlling by voltage, they are always very inefficients (losing on drivers any electrical saving gained on lighting). And as the current of the LED increases, it becomes exponentially more difficult."


Not to mention the current "creep" as the Leds "burn in".
That's not just theory either.

I'm a "cut and try" kine guy.
So, I wired a string of 5 Watters to my adjustable constant current/voltage bench supply.
Locked the current down and made a note of the voltage.
3 weeks in. the voltage had crept .2V.!

That don't sound like a lot, ya?
When were talkin' Vf on an LED it can be big trouble.
So, folks, that's good advice!
Resistors to limit current are cheap and easy.
But, resistors are linear, led junctions are NOT!
Once you get over a half an amp or so, they are a false economy.

LEDs will last a very long time when fed properly but they can not take a joke!
You can make an in-efficient but dirt cheap constant current supply from an LM317 and a sense resistor if you're on a low budget.

Got money to burn?
Buy a 120v. ac in, constant dc current out, supply for each balanced string of leds. Makes everything super easy.

O. K. who wants to play.

Show me 3 hands and I'll start an "LED Git 'er done" thread.

Aloha
Weeze

 

[Supacalabala]

"Although it is possible to run a system using a fixed voltage Power supply (PS) and resistor or other ways of controlling by voltage, they are always very inefficients (losing on drivers any electrical saving gained on lighting). And as the current of the LED increases, it becomes exponentially more difficult."


Not to mention the current "creep" as the Leds "burn in".
That's not just theory either.

I'm a "cut and try" kine guy.
So, I wired a string of 5 Watters to my adjustable constant current/voltage bench supply.
Locked the current down and made a note of the voltage.
3 weeks in. the voltage had crept .2V.!

That don't sound like a lot, ya?
When were talkin' Vf on an LED it can be big trouble.
So, folks, that's good advice!
Resistors to limit current are cheap and easy.
But, resistors are linear, led junctions are NOT!
Once you get over a half an amp or so, they are a false economy.

LEDs will last a very long time when fed properly but they can not take a joke!
You can make an in-efficient but dirt cheap constant current supply from an LM317 and a sense resistor if you're on a low budget.

Got money to burn?
Buy a 120v. ac in, constant dc current out, supply for each balanced string of leds. Makes everything super easy.

O. K. who wants to play.

Show me 3 hands and I'll start an "LED Git 'er done" thread.

Aloha
Weeze

I use power supplies for my indoor LED projects, but I think they won't have enough to power a string of 3 watters.

 

[CabGrower]

Ok, why don't we add a small 13w cfl to a red and blue led array to add the extra spectrums not covered. Try it with and without and see the difference. Then work from there. With all experiments, you want to limit the amount of variables to only those you are interested in.

 

[Weezard]

Good point!

Good point!

Ok, why don't we add a small 13w cfl to a red and blue led array to add the extra spectrums not covered. Try it with and without and see the difference. Then work from there. With all experiments, you want to limit the amount of variables to only those you are interested in.

Funny you should suggest that, CabGrower.
Added a large, (120W equivalent), CFL to the flower room last month.
Just in case the girls were gettin' sick of purple.

Did fairly well without it but would like to go from fairly well to excellent.
Be another 4 weeks before I can say if it did any good.

Peace and a plethora of pistils
Weezard

 

[Supacalabala]

I going to wire an led array to supplement my 85 CFL which is @ 5000k. The red 5mm LEDs I'm going to use are rated @ 640-645 nm.

 

[MrNebulized]

Hey Weezard, KNNA

I have been reading your guys' research on the "perfect LED light" for a long time. To bad that thread is no longer. That had some great stuff. I would get a little to toasty and have to re read but it is to bad that forum isn't up.

Still no money and hoping to build me one. I'd like to do the equivalent of a 600 watter with some extra money Uncy Sam should be giving me. KNNA you taught me a lot, I don't remember it all LOL but I did get pointed in the right direction from you. Learned a lot about Photosynthesis the light waves that require the least amount of energy to convert. For instance plants do use green light but extremely inefficiently(don't know who sad that in the thread) by converting it to a energy it can use. A lot of people don't realize plants CAN use any light, but you are looking make the most efficient use of light possible. This is why LEDs are so great because you can use the specific light NM that helps your plant uses. I remember looking at the the graphs. Showing the spike points on what NM you shoot for.

Still a lot of research to be done, you guys have been fun to watch and read about in your projects. I will still be watching as I have been for a few months. I will be asking for advice as well in the near future. I have been looking to build the most efficient garden environment for the plants, hoping to get some gnarley large buds in as short a time period as possible. Can't wait to get started on my LED set in the next month, will be seeking your guys advice once again, in the mean time I will continue to read and research!

The Nebulizer

 

[irconfused]

My yeild matches a 400W HPS with only 45W. of LEDs

And that would be how much?
Also what are your results as far as stretch and flowering times?

470's and 455's are everywhere but where did you find 430?

I'm curious as well... anyone else feel free to chime in

Aditionally, the only high power 660nm leds i know are the LEDEngin and the Edixeon Federal (that is avalaible on 740nm, far red, too), that are comparatively expensive against Luxeon, Cree, Osram or SeoulSemiconductor leds.

Who makes 640nm? I've only been able to find 620-630nm then 660s, also do you know of any place to buy the edison 660s and far reds (even though you don't recommend the 660s)

 

[Weezard]

I am Weasel!

I am Weasel!

And that would be how much???
Also what are your results as far as stretch and flowering times?

And that, would be telling.

Lotta variables to consider so, most quantitative answers to that question are meaningless.

Stretch?
That, I can answer.
How much stretch do you want?

One advantages of LEDs is almost full control.
You can increase stretch to double the height if you wish, and I sometimes do.
Or you can limit stretch to 10 or 20%.
Again, this is subject to many variables.
Genetics, temperature, root health, nutes, etc.

Flowering time? Ay, there's the rub!
Had a Sativa that took 16 weeks to finish!
Most strains do take an extra couple of weeks.
I could accelerate and add, perhaps 10% to the bud weight by tripling the power to 150W.
But, 45W. has a better R.O.I. and I'm in no hurry

If you are in a hurry, LEDs are not for you.

Once you compensate for the longer flowering time you are still way ahead on cost of power, (We pay $ .54 per kwh. here).
So I have a perpetual harvest.
Waiting for the first harvest was a bear, but now it doesn't matter.
Every 2 weeks I have a different strain ripen.
Love that variety.
Same old, gets old, ya?

So I limit my girls height to 5' including the 5 gal. bucket.
And start a new one every 2 weeks or so.

Yield?
Keeps me in meds.

Aloha,
Weezard

 

[GeneticHeirloom]

I'm pulling up a chair and throwing some K around for the knowledge Knna and weezard, keep it up!

 

[indifferent]

I'm calling BS on the 45W giving 400W results. I can easily yield over 300g with a 400, are you saying you can get over 300g with 45W of LEDs? That's a strong statement and foolish to make it unless you can back it up with some kind of proof.

Edison Opto 660 and 740nm 1W LEDs are available from this German supplier, they are bare emitters, they aren't cheap though:

http://www.led-tech.de/en/High-Power-LEDs-Edison/Edixeon-1W-c_142_145.html

Cheap red 620-625nm 1W emitters are available, they are only $1.48 each, cheaper if you buy more:

http://kaidomain.com/ProductDetails.aspx?ProductID=5528

You can get 460-470nm blue 1W emitters cheap too:

http://kaidomain.com/ProductDetails.aspx?ProductID=5525

White ones are cheap too:

http://kaidomain.com/ProductDetails.aspx?ProductId=4908

320mA drivers with a useful 42V Dc output are cheap:

http://www.dealextreme.com/details.dx/sku.13691

Those emitters aren't the most efficient but they are the cheapest option for the time being. I chose to use CREE 700mA XR-C and XR-E series LEDs:

Red CREE XR-C:

http://www.dealextreme.com/details.dx/sku.1776

Blue CREE XR-C:

http://www.dealextreme.com/details.dx/sku.1775

Cool white CREE XR-E:

P4 bin http://www.dealextreme.com/details.dx/sku.1302
R2 bin http://www.dealextreme.com/details.dx/sku.15943
Q2 bin http://www.dealextreme.com/details.dx/sku.2395
Q5 bin http://www.dealextreme.com/details.dx/sku.2394

 

[Weezard]

300g! I wish!

300g! I wish!

I'm calling BS on the 45W giving 400W results. I can easily yield over 300g with a 400, are you saying you can get over 300g with 45W of LEDs? That's a strong statement and foolish to make it unless you can back it up with some kind of proof.

Oh hell no!
I'm kind of a noob at this Led and hydro.
Been growin' off and on for 40 years and have never gottenn 300G from one bush
I'm just saying that I get the same results with 45W of LED as I do with a 400W HPS.

If you can get 300g from one plant. I am not worthy.
I only have about 3 sq ft of bloom room to work with, so.
I'm quite sure your milage and mine are not in any way compare-able.
I was growing meds with an HPS. Now I grow with LEDS.
same amount of buds, 1/10th the cost.
Well 1/8th if you factor longer time to finish.
I make no "wild claims, and I'm not sellin' nothin'.

If you firmly believe that LEDs don't grow weed.
Bless your heart, Don't use them!

I didn't know, until I tried.
Well, actually Steve Fortier's results with tomatoes and Whiskey Tango's Pot grow gave me the confidence to try.
I'm quite happy with my results
If you don't pay .54 per kwh. you have no need, ya?

Aloha,
Weezard

 

[indifferent]

Who grows one plant under a 400? That's a pretty untypical grow.

How much are you yielding with 45W of LEDs?

 

[nlvigour]

Experiments with LED's

Experiments with LED's

To date the most well documented LED grows - though not plant specific - were at

http://www.greenpinelane.com/

He had the greatest success with the TI Smartlamp & TI Smartbar from Theoreme Innovation Inc.

http://www.theoremeinnovation.com

I'm curious to know what Knna or any other knowledgeable specialist think of this TI Smartlamp. The following is the marketing literature from the manufacturer.

The TI SmartLamp can be used as a supplemental lighting for greenhouse crops or as primary lighting in closed areas.

The light integrates the ballast, the reflector and the ventilation system required for its use. No assembly is needed. Its compact design minimizes shadow effects. The TI SmartLamp is suitable for power supplies in the range of 110V to 240V (50Hz/60Hz) and is equipped with a glass that provides a high light transmittance of up to 98.5%.


Area Coverage
Recommended lamp height above ground:

• Supplementary lighting: 1.8m (6 feet) or less*

• Primary lighting: 0.6m (2 feet) or less*

*Depending on the type of plant and the desired result.


Luminous Output
Total photosynthetic photon per unit of time
183 umol/s



Electrical Characteristics
Operating Voltage
110V - 240V

Operating Frequency
50Hz - 60Hz

Typical Power Consumption
300W***

*** 330W with the TI-ProBloom option



Other Characteristics
LED estimated life span
>50,000 hours

Light output maintenance after 50,000 hours
70%

Operating luminaire orientation
lighting downwards

Maximum surface temperature
65ºC (149ºF)

Operating temperature
(Indoor use only)
-20ºC to 40 ºC (-22ºF to 122ºF)

Run-up time (100% of final value)
less than 2 seconds

Knna mentioned that a balanced white spectrum would probably be more efficient in terms of efficacy. I'd be grateful if to know if the TI Smartwhite Industrial Series would be a better alternative in the veg stage than the marketed TI Smartlamp with its redder spectrum. Or would a person get more efficient or greater photon output from a similarly sized ceramic metal halide (250 -soon to be possibly released 300 - or 400 watt?

Many thanks to those who are able to offer any insight.