DIY initial thoughts

[trunche0n]

I would be very much in favour of a full spectrum setup as opposed to dual....

However as a novice and just starting to get to grips with the theory It is easier to "half get it right" when coming up with a specific flower or veg spectrum.

I fear with a Full spectrum It would be alot more sensitive to good design and theory behind it so that it performs adequately for both cycles of veg and flower and alot more punishing for small errors in the wavelenghts used...

I will rethink my veg spectrum and reduce the CW a good bit and ww a tiny bit in favour of a bit of red............

REVISED Veg Spectrum:

42% Warm White xt-e - 8% Red 630 - 12.5 % Red 660- 37.5% Cold White xt-e ( In my vegging mode led no. = 20WW : 6(660) : 4(630): 18CW )

 

[SnotBoogie]

To be clear, my spectrum is not "full spectrum" - I simply run my flowering setup for veg and still get excellent vigorous bushy growth. I should add that i only veg for short periods - 3 weeks at the absolute most.

Chosen which 660nm and 630nm reds to use yet?

Here's another can of worms for ya: beam angle. Personally i'm a little OCD about keeping it nice and homogenous for good spectral mixing. The "advanced practitioners", though, use complementary but different beam angles to create a beautiful "cone", which is requires a little much prototyping for me. You can see stardustsailor demonstrating this in his recent post, IIRC.

 

[SupraSPL]

Check out the luxeon ES deep blue and Cree XT-E deep blue. They are both available in top bin for $3-$4 and the Luxeon is 55% efficient at 700mA. The cool white XT-E is 43% at 700mA for comparison, which is still excellent.

One thing to watch out for, high efficiency LEDs will glow from ghost voltages during lights out. I noticed a string glowing during lights out because the extension cord powering the drivers was not grounded. So I isolated the heatsinks from the wire hangers and the glowing went away. If I touched the heatsink with my hand the LEDs lit up slightly. I'm sure you can imagine how something like that could cause all kinds of problems in sensitive plants. I brought this up because someone mentioned putting the drivers on a switch. Just make sure you switch the hot wire or switch both wires and then double check to make sure they are not glowing during lights out LOL.

 

[PetFlora]

Thanks a million for the detailed replies! Cant beleive everyones so damn helpful. Thought i'd get ignored for a basic info thread. cheers

The spectrum is what is beating me at the moment - I read (and apparently didnt understand!) alot of threads here and thought my ration would be fine.

Is there at least a target spectrum of light to achieve?

maybe that way I couldsort something better out ratio wise.

I can read the data sheets ok but its knowing how much is needed of each temp is getting me ...

If you speed through my thread you will see lot of LED info that can be very helpful

 

[SnotBoogie]

Damn...never even thought of that Supra... Im gonna be paranoid now

 

[trunche0n]

Ok so I been searching and cant fin any notes/reference o ideal spacing betwen each 3w diode? Is there a magic distance? I see some people using stars mounted right along side each other.....

As for lens degree angle. I dont know.... - It doesnt seem to make sense to project on to the sides of the tent.
What is the Ideal distance of the Led Panel to the top of the plant canopy? Perhaps then I can decide whether to limit or throw the degrees

 

[PetFlora]

Let me try to unconfuse you...

Kelvin has to do with CFLs, Metal Halides, HPS, Induction and household LEDs.

2700-3000K = WW (warm white ~ 590-630)- appears yellow/red

5000-6500 = NW (neutral white ~ 450-590) appears bright white

When speaking about actinic/monochromatic leds, nanometers are used

Here's a couple shots of clones growing under a 22w NW household led bulb. 3 sets taken at different times , so height varies. First 2 pics taken on 3/26. First pic taken yesterday

 

[trunche0n]

oooooooh - (distant rumblings of realisation dawning!....)

Sobasicaly you are saying the throw of light and how we measure it are very different things between led = cfls??

am I on the right track even?

If you speed through my thread you will see lot of LED info that can be very helpful

Sorry Petflora - Yes I have much great reading from your thread and was at one stage considering the par38 merkaba setup - it does look promising.
You certainly dig that merkaba layout!!!

 

[trunche0n]

If thats the Case do I just go for top bin diodes?

For example the 3750k R3 XTE (flux 122) warm white instead of the 2700k E8 (flux 107) warm white as the flux is where its at?

 

[SupraSPL]

As you probably know the white LEDs are actually deep blue with a phosphor coating. The phosphor converts blue photons to a white spectrum but it is not a 100% efficient conversion. So the higher the color temp, the more blue that is in it and the less phosphor that has been used. Less phosphor means higher overall efficiency because it will output more blue and less white.

BUT on the other hand the lower the color temp the more efficient the spectrum is for photosynthesis (more output in the deep red range). So if you use a lot of 3750K you could add 660nm deep reds to improve the spectrum. If you go with 3000K or 2700K you could add both 630 and 660 reds and potentially more blue. KNNA recommends 15-20% blue output in flowering lamps. I ran 18% and was happy with the results.

By counting boxes under the warm white graph, I estimated that Cree XTE warm white includes about 10% deep blue output if that helps your math. In my lamps I used XTE R3 flux 7A tint which is approx 3000K. I added 630s 660s and deep blue.

I am glad to see you are planning on scrapping the all white lamp. A few years ago I built a lamp that included a higher percentage of white and was disappointed to discover it caused a higher leaf ratio similar to outdoor buds. Now I do think there is such a thing as too much white so I add 630s and 660s to the warm whites, mostly 660s. I am open to being proven wrong on this I know there are fans of the all white lamp.

Another benefit to using 630s and 660s instead of all white, they are more efficient than warm whites. The Cree XPE P3 630nm reds are 42% at 700mA. They are available, cheap and mounted on stars. The Luxeon ES EX6 660nm are 38.5% at 700mA. The XTE R3 7A warm white is 34.5% efficient at 700mA. The XM-L T3 warm white is 34% at 700mA and of course that would be twice the price of the XTE.

 

[SupraSPL]

Ok so I been searching and cant fin any notes/reference o ideal spacing betwen each 3w diode? Is there a magic distance? I see some people using stars mounted right along side each other.....

I use passive heatsinking and bare emitters so YMMV. In my case the distance ends up based on the heatsink profile. I Ccalculate the surface area of the heatsink and install 1 watt for every 16sq in of surface area (calculated for top binned LEDs). The keeps LED junction temps below my target (50c). Then just space the LEDs evenly across the surface of that heatsink. For example in veg we use a heatsink that has short fins so the LEDs are spaced further apart than they are in flowering with taller fins.

The flowering lamp with emitters running at 700mA (dissipating 2.2 watts each white/blue and 1.7 watts each red) are spaced about 1.5 inches apart center to center and the maximum distance to canopy in this case is about 10 inches. If you can give your canopy a stadium shape around the edges it would be more efficient.

 

[PetFlora]

SPL, are you running the 630 & 660 throughout v & f or are they on a separate switch just for flowering?

My thought is if 640-660 is on at all times no clear signal is sent to the plant during the flip

Edit: Even a few years ago led tech was Cromagnan.

I grew all the way to flowering with just NW (veg) then WW for flower. Only because I was growing a new breeder's seeds did I add my 3 yo ufo 90 (R/B 8:1) to give them the best I could

HiloReigns has a thread using only 28w of WW and has just harvested some nice bud

WW should have ample 630. I think we don't just want 660 (or even need), but as much of the spread in between (640-650). Anyway, 660 is oversold, probably from the ancient monochromatic peak PAR days. Reboot your mind folks

 

[trunche0n]

Thanks folks you are helping me a great deal I appreciate it.

I can get 4 of these puppies for 80 dollars shipped. Virtualy Silent and are rated for 215w TDP EACH

Idea is maybe to bond them to 4mm plates of aluminium?.....................

If I did two panels I figure I could get a way with one on each panel.

After all these are deaigned to effectively cool a more thermally challenging than most graphics card not to mention its usually trapped in a desktop case. These would be free to air so to speak and may even run passive.

 

[PetFlora]

Slap a multichip on each and, grow man, grow

 

[SnotBoogie]

Chroni your PM box is full mate

Here you go:

Morning Snot (where i am)

I just caught your post on solder paste....is this it?

http://www.aliexpress.com/item/Amtech-10CC-RMA-223-UV-BGA-PCB-Soldering-Paste-Flux-Solder-Ball-Flux-Paste-New-5pcs/537270617.html

And you mention a clothes iron? how does that work?

Think what you linked is just Flux (sticky washing up liquid for metal kinda stuff)

Here's what i got, you might find it cheaper. This is literally enough for 100's of LEDs

http://uk.farnell.com/jsp/displayProduct.jsp?sku=876768&CMP=e-2072-00001000&gross_price=true

I'm going to do a tut on the clothes iron thing, but basically you basically use the paste on the stars, plop the LEDs on them, and then put the stars on the upside-down clothes iron (so its like a flat plate (edit: i put mine inside a cooking pot, upside down, with oven gloves pushed around it to keep it stable. lawl)) and turn it on. If you get good at knowing how long the iron heats up for, you can turn it off before it peaks and replicate a "pro" reflow cycle (i don't really worry too much about it myself personally).

The cool thing about it, is that you can work with super small components that would get knocked around or jarred if you tried to do it with a soldering iron. you can also, for example, reflow a whole circuit together in strips, by laying it out on the clothes iron (what i did for WW-pontoon, which i will have photos of soon). its just soo easy. I only use the iron for stuff i cant really do on the hot plate now.

Oh another cool thing is you can get premade PCB boards for the entire circuit and all you have to do is mount the leds to it like its a giant Star.

 

[Chronikool]

Sweet....thanks...That helps heapz...i have much to learn...inbox is cleared....its hard when i get so much mail everyday...hahaha...

 

[SupraSPL]

SPL, are you running the 630 & 660 throughout v & f or are they on a separate switch just for flowering? My thought is if 640-660 is on at all times no clear signal is sent to the plant during the flip

Yes I run the same spectrum in veg and flower, just a bit more blue in veg. They have no problems switching over and finish just as quick as HPS, maybe faster.

Edit: Even a few years ago led tech was Cromagnan.

Years ago I was using Osram Golden Dragon when I determined I was getting leafy buds from too much white. The lamp was 26% efficient overall which is higher than most commercial LED lamps available today. I am sure using NW and WW will get the job done but I suspect the spectrum can be improved, especially when we have 55% efficient blues available for $3.

660 is oversold, probably from the ancient monochromatic peak PAR days. Reboot your mind folks

The Luxeon ES 660s are 38% efficient at 700mA and only $3.40. An excellent supplement to warm whites. To each his own

 

[trunche0n]

I ran 18% and was happy with the results.

Hey Supra - how do you calculate your 'overall' percantages of colour? Is it very complicated?

Concentrating on my Flowering Mode I think I want to use: (But this is a percentage/ratio of diodes NOT spectrum as they all have various spectrum values....)

45% Cree XT-E Warm White Bin Q4 2700k min cri 80 XTEAWT-00-0000-00000HCE8
25% Luxeon ES Red 660nm Bin EX6 LXM3-PD01-0300
17.5% Cree XPE 630nm Bin P3 XPeReD-L1-0000-00801
12.5% Cree XT-E Cold White Bin R5 6200k min cri=70 XTEAWT-00-0000-00000BH51

22WW : 12(660) : 8(630) : 6CW )

Do my Bin/Color Temps look ok?

I defintatley want to run absolutely everything at 700ma to keep things simple for first build.

For example if a minimum of 14% blue is needed for flowering and I am relying on Cold White to provide the blue then I must be way short as I am only using 12.5% cold white leds of which not all its spectrum is blue so I am pretty confused!!

 

[FranJan]

Hey Supra I was wondering on the above setup should trunche0n be running everything on 700? Or more to the point, will XPEs be buried by the other diodes and be ineffectual? It just seems the other diodes output will be of a certain intensity at x distance and that the 630 XPE's output will be much less and be ineffectual if they're not driven harder or even replaced. Or am I way off track here? Thanks Supra.

 

[trunche0n]

Great point Fran how about:
Oslon SSL 80 LRCP7P-JRJT (JS) RED 630nm as a replacement driven at the 700ma?