DIY LED selection for FLOWERING

[turnip brain]

Roughly 200-250W

Not multi purpose veg/flower

FLOWERING ONLY!

Having read many threads and thousands of posts, there is SO much conflicting info to sort out, The learning curve is very very steep right now for me.

I wish to simplify the design of a lamp in terms of LED selection, but in doing so don't want to leave out any important spectra for best efficiency flowering, nor end up adding something in excessive amount that would compromise overall spectrum.

Based on averaging couple people's recommendations, I came up with:

400-499nm: 12%
500-550nm: 15%
550-600nm: 13%
600-650nm: 29%
650-700nm: 30%
700-750nm: 3%

From this, the simplest arrangement seems to be all WW around 2700-3200K,

Red peak is around 610nm, and there is another blue peak near 450nm. From what I have read, seems blue is covered OK for flowering, but should I supplement with more red/far red to increase roughly the 640-700 range?

Looking at combined spectra here: http://buildmyled.com/custom-led-strip/, adding red/far red creates intense peaks in their own range relative to the overall curve of just WW. I wonder how much is too much and how to balance relative outputs if I do add red/far red to the WW.

From RRog's build thread and other discussion of the cree LEDs, seems like the xm-l's have great output/efficiency and for me, the higher wattage is desirable so I can mount/solder half the number of LEDs to achieve the desired 200-250w. Relative costs seem to be a wash here between this arrangement and running double the number of lower wattage chips like ww cree xt-e run at 700ma.

So my starting point is:

40 XM-L LEDs driven @ 1750ma x 3.2v=5.6w ea 224w total,
4- meanwell lpc-60-1750 drivers each handles 10 chips

So, all discussion/input/being pointed to resources appreciated! And especially input from empirical evidence from actual grow testing

Please tell me if I am on the right path!

Thanks

 

[xmobotx]

w/ cree neutral xml & warm xml in some combo; i dont see how you could go wrong

 

[turnip brain]

Thanks, I can see NW and CW for VEG, but why add NW at all for FLOWER only?

 

[xmobotx]

my thinking is to help keep the buds compact

cree NW is pretty close to the ideal curve for plant growth in general and has plenty of blue ~maybe try 3 warm to 1 neutral for a warm-heavy lamp

the reality is i would like to see what happens w/ 1 cool diode in the middle ~it seems to have a certain intensity which attracts the plant {like to follow the light source "sun across the sky"}

there is a lot of speculation about this ~its a pandora's box & i wouldnt be surprised if differing people got different results

 

[turnip brain]

Cool, I may be over stressing about it, but just want to understand what I am doing before building.

I have read a lot about the need for blue throughout growth cycle, but with WW there is still a blue peak in spectrum, the Cree XML spectral chart shows around 60% relative radiant power at around 450nm for the WW. Is that not enough? Same chart for NW and CW show really high blue peaks greater than any other wavelengths. Seems potentially excessive for flowering.

 

[xmobotx]

well; 'is that enough?' is really the magic question

its probably true that one could taylor the ideal lamp w/ discrete diodes but; the problem there is figuring out a) what that elusive ultimate combination is and b) now you'd have to figure what diodes voltages amperages and drivers to mix all that hodge podge together

the theory w/ white LEDs is that you get a homogenized spectrum {credit to tenthirty} w/o all that driver juggling involved w/ discretes {NTM placement}

essentially; you have a red heavy homogenized white in the warm whites a neutral homogenized white {most like sunlight} and a blue heavy homogenized white in the cool whites

you are correct that warm white alone would be fine ~my feeling is a little neutral might give some denser flowers and like i mentioned theres something interesting about the intensity of the cool white XMLs

see what i mean by "it's a pandora's box?"

 

[turnip brain]

Tempting to turn it into a hodge podge, but that is not what I am looking to do. I am really looking for simplicity.

If the use of LED lighting used to flower MJ is still in such an infant stage of development that factual and empirical evidence does not exist to answer these questions, I accept that.

If I had 1-unlimited funds, 2-unlimited time and 3- the space to do it, I'd love to run different lights under controlled conditions to find my own answers but alas, I have none of those three things!

I just want to be making informed choices.

Relating to use of WW, Really comes down to two questions:

1-Is there enough blue from WW LEDs alone?

2- Red peak from WW is at around 610nm. Should red above 620nm or so be supplemented?

Here is a link to the Cree XML spec sheet: http://www.cree.com/~/media/Files/Cree/LED%20Components%20and%20Modules/XLamp/Data%20and%20Binning/XLampXML.pdf

The thing that bothers me about the relative spectral graph is it seems not to show the WW/NW/CW spectra compared to an output baseline, but only relative to each diode's own 100% output level.

 

[xmobotx]

you are right the data set does leave us speculating

for instance; while the peak is @ ~610 for the WW; the tail of the curve goes on to >750 so; the question becomes 'is it enough?' & the mechanical ramifications of supplementing it move away from simplicity {esp w/ questions like 'how much' & 'where'}

i wish i {or someone} could give definitive answers ~i have read a ton of knna's and sailor's data which just gives us more to speculate about and ends up being debated w/ no real answers {hard to sort through what the answers may be anyway}

what we need is some real-world testing and hard data not anecdotes and opinions tempered w/ speculation and confirmation bias

anyway; people have had successful flowerings w/ combinations of WW and NW and/or CW in varying ratios ~hard to say whose was 'optimum'

honestly the lamp i am most impressed w/ is trem0lo's KNNA-based combo of ww,660,&RB but; those LEDs are osram or something and not even the best available by the data {dated/obsolete? ~lol}

i m not seeing the same kinds of results attempting to emulate it so; much of that credit must go to trem0lo

 

[turnip brain]

I'll try to look that up. I've read so much it is a pretty big blur and don't remember that particular project.

 

[xmobotx]

idk why i didnt provide the links for that ~it may not be like everyone just knows what a 'trem0lo lamp' is {but it should be}

this is the original thread

here's a second run/same light

not sure how that may translate to a larger grow ~as you can see that is a PC grow in the micro forum

 

[PetFlora]

From extensive reading of DIYers and A51...

For flower only ~ 8% blue either CW or NW, cause you need some (~20% for veg).

Quality bin WWs should provide good spectrum through 640

Any additional red (650-660+ < 10% total)

730-740 on a separate o/o switch ~ 5%

 

[turnip brain]

Thanks both!

Pet flora, in addition to reading DIY info, I have been looking around LED grow light fixture manufacturers to see whether they share spectrum/emitter info. A-51 unlike most fixture manufacturers, does this. I like their no BS attitude. Would like to compare to other commercially made fixtures if you know any others that share this info.

Do you consider A-51 a benchmark in some way? Kind of seems so as they are mentioned pretty frequently by DIY ers

 

[jcmjrt]

A51 does make one of the better commercial lights. Also take a look at Hans panel at bonsaihero.com

 

[turnip brain]

Hans- 4 Royal Blue XT-E, 27 Red XP-E 501 and 2 DeepRed Rebel Led's 56-65w

A-51 48x Cree XPG Outdoor White + 24x Cree XPE Red 155w

Debate on R/B vs supplemented whites...

 

[turnip brain]

BTW, looking at commercial fixtures only as reference. Mine will be DIY!

 

[xmobotx]

i think CW XMLs supplemented by 660 red might just be a raging combination ~the question becomes 'what ratio?'

it would seem a little lacking in near red but plenty of lamps do pretty good w/ RB/deepR and the CW/deepR combo would be a step towards homogenized light from that platform

 

[turnip brain]

Well, this is not calculated yet, and I have not found/received the answer whether WW provides enough blue, but I am starting to think about adding some CW to help bump blue a bit anyway. Certainly better to have some more in early stage flower, but not sure in terms of overall averages throughout flowering cycle without adding dimming etc.

So would still be a base of WW, supplemented by some CW, and some deep/far red. SDS argues a lot against the 660nm though, and leaves me wondering if the WW red 620 or so peak will suffice. not settled on that issue. still need to find more info about red range.

Also now considering better efficiency and spread/mix with XPEs, but not sure if I want to mount/solder double the LEDs.

 

[Gru]

Maybe these articles will help you.

http://www.lzi.lt/tomai/97(2)tomas/97_2_tomas_str11.pdf

 

[xmobotx]

the nasa tests found that very little blue was needed to grow like lettuce or something

of course we also know that a little more blue light tends to keep canna a little more compact

YMMV on how that affects your decisions and what is optimum for 1 strain may not be optimum for another as well as what is optimum for 1 grower may not be optimum for another

the only way to definitively answer your questions would be for someone to do side by sides w/ that various configs of interest @ the scale you intend to grow

i did wonder about and read in to the discussion about too much 660 and possible adverse effects/benefits of more near red ~its basically another refer to the above scenario {strain/grower/scale dependent and the evidence/data we are working from is anecdotal/subjective}

so a kind of thing where; yeah ~ WW will work by itself/a little more blue wouldnt hurt {particularly if you may prefer more compact growth} and some more red probably wouldnt hurt either {but then you get away from 'simple'}

there are those having success w/ some various differing combos of WW/NW/CW & the consensus is that XML defo delivers as far as lumens/watt and intensity

and; its a wormhole where you can just keep considering varying possibilities, combos and, anecdotal data sets ~probably best to look for data sets close to the scale of grow you want to pull off

this thread has a lot of consideration relative to white cree diodes

 

[PetFlora]

A51 is very tuned into developing its' panels for mmj. Been following their posts in DIY on another site- very helpful, informative, with vg 411

Hans as a vg panel for closet grows, but has been very stubborn about using white in place of R/Bs

xmobotx

I am in agreement regarding CW (~6000K). Alas, CW not common in household globes stuck with NW (~5000K). That said, their actual colors are all over the place. I have several 5000K that look more like 6000K

IMO, 660s are oversold, but probably need more when the whites are CW. >90% of PS takes place in the ~630+ nm range. NW have ample ~ 630, and therefor do not require much 660, say ~ 10%, whereas CW needs ~ 10% 660, but also need 640 & 650 (~30/20%)