"Home made" LED grow lights...

MajorCoco

Well-Known Member
I'm trying out a grow experiment, since my trusty 400W HPS was in storage and hard to get at. I've always thought LED was the way to go, so I thought I'd splash a little cash on some reckless experimentation and share the results as I go.

I'm just doing a very small cupboard grow, so the aim was to get about 250W HPS equivalent of lighting.
As a starter, I had a 65W 2700K CFL and an 85W 6400K CFL. These were to keep me going until I'd ordered my parts. Parts for this were easy to find. I'm in the UK, and these are all available on the internet easily. I bought:
1 x Bridgelux 9000lm 5600K Array LED (£60)
1 x Bridgelux 7000lm 3500K Array LED (£60)
2 x LED Driver (100W each, 2.8A) (£70 each)
2 x Heatsinks (Big ones!) (£20 each)
So, not a cheap experiment! £300 for the key ingredients.

The heatsinks were designed for these arrays, so once I'd soldered the leads, it was pretty easy to screw the arrays on to the heatsink, making sure to use some good thermal paste. The arrays were tricky to solder with my puny iron, but if you've soldered before it shouldn't be too challenging.

IMG_1112a.JPG
Here they are hanging up. The heatsinks are about 20cm long, and are painted aluminium. Chunky!

There are a few reasons I think LEDs are better than HIDs.
1. Directionality - LEDs put all there lumens in one direction. (These arrays say about 120 degrees angle of output) HID lumen output is measured all around the bulb, some of which you lose in the reflector. With LEDs, this automatically gives them an efficiency advantage as up to 20% of light leaving the bulb going up will be lost. (So a 50,000 bulb actually could be sending only 45,000 lumens down) All the lumens from the LED are heading down, and even without a reflector there's very little coming out sideways.
2. Spectrum - HIDs put out a lot of energy in infra-red (IR) frequencies. These aren't useful for growth, and actually end up burning the plant if they get too close. LEDs of this type use blue LEDs with a phosphur coating to provide the wider spectrum light. There is very little energy wasted in the IR spectrum.
Here's the LED's spectrum:
Capture.jpg

Both have sharp peaks at 450nm and a wider peak from 550 to about 675nm. Unsurprisingly, the difference is that the red array puts out significantly more in the red end, and the blue puts out more at 450nm. These aren't far off ideal, although slightly more at 400nm in the blue, and more up to 700nm in the red would be better. Still..less UV is probably not a bad thing anyway, and these peaks are well within the plants preferred light frequency range. I know some people get very in depth when talking about PAR, but from what I can see these lights are equal or better than HID systems which we all know work fine despite their patchy spectra...so I'm pretty certain they'll work just fine. Intensity is just as important as colour as long as the light is in the right sort of ballpark.
3. Long-lasting - The lifetime is about a thousand times longer than an HID bulb and is pretty sturdy compared to a glass bulb. Nuff said.
4. No ballast? -Well, the LED drivers are balasts, but they're pretty compact. They also put out low voltage, so they're much safer in a grow room. You could probably still get a shock though given there's almost 3A. I wouldn't like to test it too thoroughly.

Based on the manufacturer estimates, my setup is running at 320W, and is outputting 25,800 Lumens. (81 Lm/W...not bad) As mentioned above, more of this light is going in the right direction, and less of it is being wasted in IR frequencies. This means that I should easily have as much useful light as a 250W HPS system. A 250 HPS would be much cheaper of course (at least to start with), but where's the adventure in that? I also believe that a HPS would be less useful in my small (and short!) grow space than LEDs. I could have done this much cheaper by building the LED driver myself. That would have save a third at least. If I could have salvaged or DIY'd a heatsink I'd have also saved a bit. Ideas on building a cheap 100W current-control LED driver would be welcome.

Grow-
2 White Ice fem - White Label
All organic. Soil mix: 20% peat, 30% perlite, 50% wormcastings (not a good mix, but I didn't realise just how high quality the wormcastings I'm using were until after I'd first watered. They are very fine, and this has caused major problems with slow development)
From seed. Easy germ, but slow since then, solely due to an initial overwatering of the worm-casting heavy soil which I'm using. Even generous handfulls of perlite in the mix didn't improve drainage enough to help get enough oxygen in there, and the only thing to do really was to keep things warm and dry and be patient while the soil dried...unfortunately I wasn't patient, and tried to re-pot one with predictably disastrous results. The roots were tiny and fragile, as expected in oxygen-poor starter soil, and I ended up breaking off almost half of the tap root. I should know better than this. I though I was past being over-attentive by now. Re-potted it anyway, and all seems ok now. Its just starting to grow a little stronger today. Finally the roots are getting some air!

Lights look good so far. I need to run a fan over them when on otherwise the heatsink hits 60-70C, but a small computer fan at the top of each would do the job perfectly. At the moment I'm just running a desk fan for general ventilation and cooling. The heatsink sits at 40C with the fan on, giving me no problems with heat.

Day: 25C Night: 19C
Humid: 45-50 (on the low side I know, but I'm trying to sort out my damp soil problem)
Lights: 18/6

I'll post pictures comparing the brightness of the CFL vs the LEDs shortly, and let people know how the plants like them when I'm sure. They seem to be doing just fine now, slow start aside.

CC

Edit:
Here are some comparison shots, taken with identical aperture and exposure:
I know I should use reflectors on the CFL for a fairer comparison, but this setup helps to show why a flat light emitter like an LED makes having a reflector not quite so vital! I also don't have reflectors for the CFLs, which is probably the main reason!

65W Blue and 85W red CFL: (Total 150W)
IMG_1112.jpg
85W Red and 85W Blue LED: (Total 170W)
IMG_1111.jpg

Overall I'm pretty happy, and should be able to grow to within 10cm of the lights if I have to without worrying I'm going to char my flowers...
FYI, I'm running all four lights, which looks like:
IMG_1110.jpg
 

MajorCoco

Well-Known Member
Here's the set-up:
IMG_1117.jpg

Very basic. You can tell it's been years since I last did this too, as this pair are tiny for 2 weeks (and technically it's 3, since they had a week just under CFLs. Embarrassing progress I know, all because of a bad soil mix)
Anyway, I've got a much lighter soil mix seedling in the back which is 1 week old. As well as the better soil mix it's in a small pot, so I can definitely ensure it isn't over-watered, and see how it compares. I hope it grows faster!!

Here are some close ups:
IMG_1120.jpgIMG_1118.jpgIMG_1119.jpg
 

PSUAGRO.

Well-Known Member
Yep. Behind the yellow resin there's an array of about 20 smaller surface-mount LEDs...so it isn't too dissimilar to your average Cree. Just neater packaging.
Awesome little diy project you got here coco......how your "array" handle moisture(overspray/splash)???
 

MajorCoco

Well-Known Member
Cheers everyone. Nodes are looking amazingly tight at the moment, 4 nodes in only 4cm. I'm still suffering from the early overwatering problem, which seems to be causing some early nute issues, so I might have to get on with some foliar feeding.

Awesome little diy project you got here coco......how your "array" handle moisture(overspray/splash)???
It shouldn't be a problem. The resin on these can be cleaned using iso, which I've done, so it isn't water soluble, and is very probably splash proof. I've left it without cover for simplicity, but it wouldn't be too hard to get a polycarb lens on it if you wanted to protect it a little better.
 

MajorCoco

Well-Known Member
Cheers. I was getting a bit worried, because growth is still too slow, and I've got some discolouration on older leaves, so I decided to go ahead with some light watering with ph6.8 water to try to flush out any early ph problems caused by my ph8+ tap water. I also double potted one, and covered another with newspaper in case light was getting to the roots. I'm trying anything I can think of now!!

Here are more recent photos:
View attachment 2089440
View attachment 2089442
View attachment 2089443
They're still too small to flower. Another 2 weeks at the very least I think.

Does anyone have any ideas on the yellow and purple discolouration in the last pic? I guessed it was just lock-out due to my bad starting conditions, but I'm not really sure.

I'm not too stressed yet though as they still seem to be growing solidly, just very very slowly.
 

Attachments

Early pH issues + overwatering = no good.

http://www.greenmanspage.com/guides/plant_abuse.html

I'd wait until they dry out a bit, then flush 'em again for good measure, cut back on the foliar feeding for at least a week and see how they do, and use RO water from here on out. Tap's way too high.


Also - how high above the plants are those DIYs? Shouldn't have 'em any closer than 30cm at this point.


Cheers,

-TL
 

MajorCoco

Well-Known Member
Cheers TL.
I'm not too sure exactly what was going on, and I think now that I was a bit wrong about having an overwatering problem. There was definitely some nute lock-out, but it was pretty localised and very slow developing. Do you know if specific areas of the root tend to be linked to specific branches? Do the lower branches tend to feed from the earliest root growth for example?

I say that, because those are the parts of my pot that have been bone-dry for a week or two now, so maybe they were being kept going by the parts of the root ball which were in the damp zones? Just a theory...It's probably just because these things always show on older leaves, but I'm just wondering why...

Anyway, I gave them both a good watering yesterday morning, and they've rewarded me by instantly perking up. :) They've grown at least 1cm already today. Good healthy coloured growth too. I wish I'd watered sooner!! I guess I mistook the curled down leaves for over watering...which in hindsight I think was wrong. A good lesson for my next grow, which will also be worm casting based! Wish I asked for advice sooner though....I might have flushed a week ago and be thinking about changing to 12/12 by now. I'll be sure to ask sooner next time guys.

I've moved the LEDs up a little (now about 30cm above the tops), but at almost a month old I didn't think I needed to be too careful with light intensity? I'm probably getting about as many lumens as a 250W HPS, but without the heat. I can actually leave my hand 10cm under the LEDs comfortably without heating.

Let me know on that one.

I'll give them a bit longer this evening then put up some post-watering pics....a bit of a before and after :clap:

C
 

MajorCoco

Well-Known Member
Here's the latest update: Week 4.

The left one is the older more Sativa-like phen that got it's roots trashed at week 2. The bigger one is only just 4 weeks old, almost 8" tall, with the 8th node emerging.
 

Attachments

IlovePlants

Well-Known Member
Interesting grow major! Can't wait to see the results. I'm sure that the satisfaction of a success would be great for you considering you built the lights. Glad that some one is using arrays on the led forum for once, aside from the occasional kessil run. From what I've seen arrays have something that led panels miss: penetration. Have a wonderful grow, can't wait to see the results.
Sincerely,
ILove
P.S. Subbed
 
Hi Major. Do you regularly test the pH of either your soil or your runoff? If not, might be a good idea to do so.

Do you know if specific areas of the root tend to be linked to specific branches? Do the lower branches tend to feed from the earliest root growth for example?

I say that, because those are the parts of my pot that have been bone-dry for a week or two now, so maybe they were being kept going by the parts of the root ball which were in the damp zones? Just a theory...It's probably just because these things always show on older leaves, but I'm just wondering why...
No, the roots are merely the vessels, they feed the entire plant. Still better to have a healthy root mass throughout though, of course. Problems that tend to show on older leaves are generally due to deficiencies (either from a lack of, or from lockup) of the mobile nutrients, like Mg. Note with a high pH, problems like Mg and Zinc lockout (among others) are pretty common.

I'd flush 'em again if I were you... and test your runoff!

I've moved the LEDs up a little (now about 30cm above the tops), but at almost a month old I didn't think I needed to be too careful with light intensity? I'm probably getting about as many lumens as a 250W HPS, but without the heat. I can actually leave my hand 10cm under the LEDs comfortably without heating.

Let me know on that one.
It's probably worth remembering that both Oversaturation and Photoinhibition from Phototoxicity happens much more easily with LEDs. My personal opinion: anytime you've got more than ~40w/s.f. of (average, consistent, well spread-out) coverage with LEDs, it's too much. Reduce it.

See those leaves? Rather curled up, and pretty thick, aren't they? My guess is their thylakoids are turned onto their sides from the standard dynamic twisting they create to avoid (prevent) their greatest surface area from their reaction centers being exposed towards the light when there's too much of it. More ain't necessarily better, unfortunately.

LEDs generally give you darker leaves, too, unless they're too close. With high watt HPS, the plant creates rather inefficient, light green, thick leaves ('sun leaves'), produce less chlorophyll, and hide 'em deeper into the leaf, to prevent overstimulation of the reaction centers, esp. at full sun.

Darker ('shade') leaves are more efficient, have more chlorophyll, are thinner, and work better indoors under LED at moderate light levels. The plant will also produce less 'sunscreen' (Anthocyanin et. al.) since it doesn't need it, and put more energy into flower production instead.

That's one reason why a consistent LED array providing ~400-700 umol PPFD is a much better, more efficient alternative indoors, too...and why it's usually better to use more, lower wattage emitters spread out over a larger area, than trying to push more light out of fewer, high-watt LED arrays, which in effect reproduces many of the negatives we get from traditional point source lighting (and which we're trying to get away from).

-----

I'd try 1) raising the lights up higher/attenuating their radiance, and/or 2) looking at reducing the mA to your circuit. Are your drivers dimmable?

Also, do you still have those CFLs in there? Don't need 'em!


Just my $0.02. :-|


Cheers,

-TL
 

MajorCoco

Well-Known Member
Hey. Cheers for that info. Very useful stuff. One follow-up question though; is the problem with too much irradiation from LEDs perhaps because they (grow systems) tend to use very narrow band LEDs? Photosynthesis occurs over a wide bandwidth, so perhaps the issues you mention are due to over saturation of those fairly narrow frequencies? I'm not trying to argue you understand. It is a genuine question I have had for a while about using such narrow band emitter grow systems. I have looked at how photosynthesis works from a physics angle (it's what I studied at university), and I did wonder if a wideband light-source might result in more efficient photosynthesis..allowing more umol of light to be used?

That's actually part of the reason why I chose to do this experiment with these particular LEDs :) I like the way that phosphor smears the spectrum into something a little closer to actual sunlight than the multiband systems...albeit with some loss of efficiency.

Don't get me wrong. I get what you're saying...that my experiment is looking like failing because I'm using too much light and stunting growth! But I thought that 25000 lumens was pretty average for my 3 ft2 grow? About 8300 lumens max per ft, though I'm unlikely to be getting anywhere near that, as my temporary grow set-up wastes quite a bit of light through various gaps in reflectors. It could only be 5000, but I'm just guessing really.

I've removed the CFLs as you suggest anyway and see how they react after a week...My main concern is that the biggest is so densely bushy that it's hard to get any light at all on the lower nodes' growing heads without the CFLs. Though I may add them back at flowering time if penetration becomes a major problem.

On a lighter note..I noticed the first few female pre-flowers hairs today on both of the larger plants. Showing up in several different sites on the larger midget... but only two hairs yet on the micro-midget! Here are the pics (sorry the first isn't too clear, but it's hard to get a lens into all that foliage!!) :
IMG_1153a.jpgIMG_1154a.jpg

On your question about Ph, I did a test about 2 weeks in, which gave a ph7.1 for the runoff of ph7.0 test water. I've been running ph6.8 water since then. I've been so light on nutes I didn't think the ph would have changed much, but I should check. They're still very wet from their transplant yesterday, so I'll have to wait a bit for that.
I do think in hindsight that I've been way too light on the nutes. I'm slowly upping it with stronger worm tea plus extra goodies for an N boost. Plus some seaweed extract and micronutrient feed (no molasses as yet...I'm on the case though). The bushiest one is obviously a squat, bushy indica dom, so I'd expect darker leaves, so we'll see if less light and more N make a difference.

More updates on the way later this week..
 
Top