LED Custom 140

I have been working at this for awhile and finally got it lit up today! (Thanks to KNNA for bringing me up to speed and sourcing rare parts!) $425 for materials. Using Golden Dragon Plus premium bins. 64 reds (640nm), 14 cool whites and 7 blues (450nm). Tried to get the whites and blues spread evenly as possible. Used a ton of heatsink for passive cooling and longevity (150cm2+ per watt/ 40,000+ hours with 90% lumen maintenance) I got the heatsinks on eBay for $100 cut and shipped. These are the same as the ones from Heatsink USA llc. Going to need (4) drivers to get the job done on 120v line. Since I only have 2 drivers I'm using half the modules for now.




Added the Kapton layer and copper traces. One string for red the other for blue and white. Each string has its own driver with adjustable current. Got the Kapton and copper tape from eBay.


Thermal path is LED slug->Arctic Silver Ceramique->expanded copper tape pad->Kapton Layer->Heatsink. I did the soldering outside in the frigid deep winter. Luckily the solder I had was lead/silver and melted very fast even with an old 30 watt iron. Once I had my technique down each led only took a few seconds. I drilled a few small holes in the fins and zip tied the leads to relieve stress on the solder points. I also drilled holes for the chains to hang them.


I am not using lenses or reflectors. These modules will be hung about 6" from the canopy and the plants are LSTd so will not grow taller than ~15" or so above the soil. These suckers are so bright I prefer not to look at it even from off angle. Looking into these emitters directly would be a very bad idea. Without flash:


With flash:


Update: I finally got a new multimeter and set the driver levels.
*The LEDs will dissipate 143.2 watts.
*Input power will be 162 watts.
*Average driver efficiency is 88.4%.
*Each red will dissipate 1.4 watts @600mA and Vf of 2.3.
*Each white/blue will dissipate 2.4 watts @700mA and Vf of 3.5.

I would have liked to add some 660nm but as far as I know there are none available on the AlInGaP chips yet which are 2X more efficient. Update: These are coming soon.

I can't believe how much light these are putting out, especially when I consider from a plant's point of view it is even brighter still.

I'm still not sure what I think of leds, but this is a very nice looking build. Good job supra.

Thanks Orygun. I have the humble expectation of 1gr/watt (~150grams) for at least 10 years from this lamp depending on strain. Some have reported as high as 1.8 gr/watt.

The way I understand it, these lamps should beat HPS full spectrum systems by ~40%. In other words to outperform 600 watt HPS Hortilux you would need 360 watts of top binned LED. My lamp cost about ~$425, so that would equate to $1100 to replace a proper 600W HPS with a passive cooled top bin LED at current prices (and that is if you have a hook up for top binned LEDs)

There is also the issue of penetration. I grow LST and keep things hella short, so I don't need lenses. Also, electrical consumption is one of my primary limitations. Because of these issues LED is more suitable to me that HPS, but that may not be the case for many others grow styles.

Time to start testing it out. Running half the modules over approx 2 sq ft area. Running some Qleaner and Durban Poison. They were hanging out under a pair of old 70w HPS and they experienced some nasty temp swings due to some winter mishaps. So I won't be judging these girls but hopefully the next round.

deleted post

Now that's impressive! I was thinking of constructing my own LED system but what I had in mind was just supplementary/side lighting to overcome the low penetration problems I'll have with my LED panels once it's all up and running. You seem to have bigger and better things in mind for that little beauty. Well done indeed. There was a time when I thought a homemade cool tube was ingenuity/improvisation at it's best when it came to lighting issues. This is something else altogether... I want one..! even though my needs are far less than what your light would provide. I never thought LED technology would take off actually. I based that on an ignorance of the efficiency and benefits LED's offer and just thought HID would be the only reliable option. Man thing's have changed quickly and anything that eats less juice and gives results is a step in the right direction... Thanks for sharing your work, you've impressed me enough to seriously look into doing something similar...

Hi ASA. I'm using a pair of constant current AC-DC converters, one for each string. They are getting a 120v input, but it would be able to drive twice as many leds per driver and might even be more efficient with a 240v input. Unfortunately my living situation doesn't allow for me to isolate a 240v line for it, but someday down the road I'd like to give it a try.

Thanks Mountain Monkey, glad to see more people considering LED to save watts. Supplementary lighting is a great idea. I wonder if you could use a flat aluminum scrap panel without fins and just spread the LEDs out quite a bit?

If you imagine a supplementary panel used for side lighting, one issue that comes to mind is how to arrange the colors? If you placed 9 reds 2 whites and 1 blue and they were spread relatively far apart, while also being very close to side side of the canopy, I assume they would lose some of their spectral efficiency. I have been trying to sort this issue out in regard to the vegetative lamp design as well. One idea I have been considering is to use Golden Dragon plus 640nm red and place it near Golden Dragon plus Ultra White, then continue that pattern. The ultra white provides a nice peak just under 450nm, which would compliment the red perfectly and also would provide a nice spread of frequencies from the white phosphor.

I am planning to start replacing veg fluoros with LEDs in this style. Hopefully soon I will get a chance to test this design.

33 watt LED Garbage pail moms

33 watt LED Garbage pail moms

I scored a decent 700mA driver on eBay for $15 so I decided to try one of the modules for vegging (although it was not designed for this task). I bypassed a few of the leds on the string and jumped the strings so that the temporary driver could feed power to both the red and white/blue string. The circuit I settled on measures 38volts and 730mA so the LEDs are dissipating 28 watts. It's getting 120v AC and is drawing 33watts/38VA, so the driver is 85% efficient and the power factor .86. If I push it past 35watts/40VA it starts to flicker, so I backed off a bit and it's working great. The reds are being driven harder than I would prefer, but this is just a temporary experiment and a great chance for learning and comparison. Hopefully soon I can put together a proper LED vegging lamp.

Now I can ditch those 50 lumen/watt R20 fluoros

deleted post

hey Supra... where do you get your LEDs from? also i cant seem to find any Osrams in 640nm flavor, all i can find is 625nm

very nice thread I'm glad to see a good DIY LED panel finally. Keep it up and soon many more growers may try this as the price comes down even further.

SupraSPL said:

Thanks Orygun. I have the humble expectation of 1gr/watt (~150grams) for at least 10 years from this lamp depending on strain. Some have reported as high as 1.8 gr/watt.

Click to expand...

1 gpw is not humble That is a goal that few growers reach, no matter what light you're using. It is much more about technique and knowing your strain. Just because some have hit it, and do consistently, doesn't mean it is the norm by far. Don't get me wrong, I'm sure over the course of 10 years you'll hit it. But I hope you're not expecting it soon, and aren't the type to get discouraged when you don't hit it immediately.

I'm curious as to your electrical consumption issues. What is the main concern there?

Love the lights though man, awesome job on the DIY. I'm curious how much heat those actually put off? Just noticing the size of those sinks, looks like they could cool a Buick Do they kick off enough heat that they're warm to the touch, or hot enough you can't touch for more than a few seconds?

Again, great to see a DIY LED project going

friggin awesome job!

@ ASA. Nice LEDs you have going! I will be taking a closer look at your sig links. The drivers came from KNNA so you might already be familiar with them. They accept 120v or 240v AC input. They have an adjustable potentiometer on board. The pf changes depending on the load but generally it is somewhat low ~.6. The driver I got from eBay accepts 120v AC and puts out 700mA. It can support up to 48v DC output. Unfortunately there was only one on eBay but I will keep my eye out for any others. Where do you get your drivers from?

@ahhhavalanche! The LEDs also came from KNNA. It is difficult to get the desired bins from OSRAM because they sell them by groups of bins. KNNA must have found a way to get the value point LEDs in bulk. Here is the info I found for these LEDs (I think the peak nm band shifts to longer wavelengths when it comes up to temperature, especially the reds)

-Red bin JX 640nm (638-642nm peak)
@ 400mA, ?v, 45-52 lm, 1.13W
@ 600mA, 2.34V, 1.63µE, 1.4W (1.16 µE/W)
@ 700mA, 2.44V, 1.69µE, 1.708W (0.99µE/W)

-Coolwhite bin KZ Tone 6Q (6200-6400K)
@ 350mA, ?v, 97-112lm
@ 700mA, 3.46V, 2.05µE, 2.422W (0.85µE/W)

-Royal Blue bin 1U 450nm (449-453nm peak)
@ 350mA, ?v, 450-500mW
@ 700mA, 3.42V, 2.6µE, 2.395W (1.09 µE/W)

@junior grower thanks! I hope to see more of these get put into use. They sure are great for vegging and for micro grows even at today's prices. Hopefully KNNA will get his online store going. I think he is planning to offer complete lamps as well as parts for DIY LED builders.

@Nitetiger THNX! I agree 1 gr/w is a sweet deal when we are referring to HPS growing. My first HPS+LST grow I got lucky and did very well, but it was hard to measure gr/watt because I used 3 diff strains and they finished at very different times, which really skews the results (plus I probably didn't know how to trim properly either) Many times a plant is only half harvested and the rest harvested later on. So I never really get a good gauge on it, but I will give it my best try this time.

KNNA Quote: "As reference, other people using LED lamps using my components are getting anywhere between 0.8 and 1.8g/W, depending of grower's skills and strains used, of well manicured dry bud. Most of them are getting well over 1g/W. You can calculate your future yield by multiplying LED Watts installed by 1.2g/W as a good and conservative reference, on a 2 months basis. Adjust it according to how good yielders are the strains you like."

He also claims that you can match yields by replacing 100 watts of HPS with 60 watts of LED (properly constructed with top bins). He uses photon counts, PAR efficiency and experimentation to get these numbers. Therefore if a good grower can get 1 gr/watt with a 600w HPS using a given strain, he or she should be able to (after some adjustments) achieve 1.6 gr/watt, even when using a tiny LED.

Anyway that is my hope and the explanation of why I claimed 1 gr/watt to be a humble expectation. That's where the romance is (for me) that I could potentially match or beat 600w HPS results in a mini grow setting (gr/watt basis). If I fail to meet that, I will not give up. I can see so many places where there is room for improvement in my process and I enjoy the challenge. Due to my experimenting, I have been hit by a ton of problems and I rarely get a proper yield (but I am closing in on it ) In reality, my primary concern is quality and strain hunting. Once I settle on a few winners, then I can do some solid testing on yield.

In regard to electrical consumption, the next most efficient option is a 600w HPS. With a good digital ballast that probably draws 660w. If you are using a cooltube/cooling fans + good size carbon filter you have to power a pair of good fans (approx 100w each?) The electric cost for running a system like that is approx $46/month (+ $15 for veg costs). Granted that is not a terrible cost and would yield a lot, but I am after personal use and experimentation. Assuming I use two modules (85w)+ one for veg (35w) the monthly electric cost would be $8.00 per month. If I could get a QP of top quality dank every 2 months for $16 that would really be something!

You are correct, those heatsinks are kinda big! They are a little over 5 pounds each and measure 10"w X 6"d X 2.5"h. They stay close to ambient temperature, which I try to keep at ~75-80f (24-26c), so they are barely warm to the touch. For an idea of how much actual heat leaks into a grow chamber from 2 of these modules (75 watts dissipation), if you could imagine a 40 watt heater element (a tiny crockpot) that is about it.

There is no doubt that this much heatsink is overkill, but it is designed to maximize lumen maintenance to >90% for 10 years (or more!). It is also intended to do that job with passive cooling only, even in summer. The range that KNNA recommends for passive cooling is 125-150cm2/watt of dissipation. With this setup I am using close to 150cm2. As LED radiometric efficiency improves, less and less heatsink will be needed.

@Phillthy THNKS! What kind of OG do you have going? I will check in on your links. Tried to clone an OG Abusive and it is giving me a hard time. Alive but not rooting after 3+ weeks.

A leafy chunk of Qleaner. This one was the runt from a pair of twins that came out of a seed. They were twisted together so I separated them and I'm glad I did because it seems to be a quick finisher. The smell from the trichs is amazing. Lemony Grape in a big way. When smoked turns into a mouthful of lemons and smokes very smooth. This was grown under 70w HPS, it only spent 1 week under the LED. I am hoping for much better results with LED bud.


This a diff Qleaner lady that is at week 4-5 (LSTd with 15" canopy) She was a twin also but was not the runt. Her twin turned out to be a male. Another strange thing about this one is that it started off with a whorled phyllotaxy (triple nodes) but seems to have grown out of it. It is just starting to ice up and has the same lemony-grape resin.

SupraSPL said:

@Nitetiger
KNNA Quote: "As reference, other people using LED lamps using my components are getting anywhere between 0.8 and 1.8g/W, depending of grower's skills and strains used, of well manicured dry bud. Most of them are getting well over 1g/W. You can calculate your future yield by multiplying LED Watts installed by 1.2g/W as a good and conservative reference, on a 2 months basis. Adjust it according to how good yielders are the strains you like."

He also claims that you can match yields by replacing 100 watts of HPS with 60 watts of LED (properly constructed with top bins). He uses photon counts, PAR efficiency and experimentation to get these numbers. Therefore if a good grower can get 1 gr/watt with a 600w HPS using a given strain, he or she should be able to (after some adjustments) achieve 1.6 gr/watt, even when using a tiny LED.

Click to expand...

I actually wasn't taking into account that you'll trip that golden ratio a lot quicker with LED than HID.

Do you have any links to the grows where guys are getting the 1.8? I've been watching LED tech for quite a while, and have yet to see anything that makes them worth the cost to me compared to HID. I'd love to see a test grow that got some great results. If you have any that show an LED getting the same yield as its HID equivalent, that's the one I really want

If I fail to meet that, I will not give up.

Click to expand...

Good deal, that's what I was worried about. A lot of folks who come in with particular yield numbers in mind get very discouraged if they don't hit their target. Just makin' sure you knew what was up, you obviously do

In regard to electrical consumption, the next most efficient option is a 600w HPS. With a good digital ballast that probably draws 660w. If you are using a cooltube/cooling fans + good size carbon filter you have to power a pair of good fans (approx 100w each?) The electric cost for running a system like that is approx $46/month (+ $15 for veg costs). Granted that is not a terrible cost and would yield a lot, but I am after personal use and experimentation. Assuming I use two modules (85w)+ one for veg (35w) the monthly electric cost would be $8.00 per month. If I could get a QP of top quality dank every 2 months for $16 that would really be something!

Click to expand...

Well, first point is just me nit-picking, but you don't need 2 fans

Second, your costs aren't just electric. Make sure you also figure the time you spend into the cost equation - don't sell yourself short!

You are correct, those heatsinks are kinda big! They are a little over 5 pounds each and measure 10"w X 6"d X 2.5"h. They stay close to ambient temperature, which I try to keep at ~75-80f (24-26c), so they are barely warm to the touch. For an idea of how much actual heat leaks into a grow chamber from 2 of these modules (75 watts dissipation), if you could imagine a 40 watt heater element (a tiny crockpot) that is about it.

There is no doubt that this much heatsink is overkill, but it is designed to maximize lumen maintenance to >90% for 10 years (or more!). It is also intended to do that job with passive cooling only, even in summer. The range that KNNA recommends for passive cooling is 125-150cm2/watt of dissipation. With this setup I am using close to 150cm2. As LED radiometric efficiency improves, less and less heatsink will be needed.

Click to expand...

Bad ass

Nice This is awesome.!

@Nitetiger, I love questions that make me think and some of the statements I made in my original post definitely require qualification. You make a great point regarding time. Building an LED isn't exactly a quick project. Even planning the layout of the bulbs and the traces took me hours the first time. Actually laying the traces themselves took almost 6 hours and nearly killed me from breathing fumes from the conductive acrylic adhesive. Affixing the LEDs and soldering them took a few more hours, working out the method of safely mounting the connecting wires between modules, wiring for input power for drivers, switches and drilling holes for mounting chains took a ton of time also.

Then there is the time I spent reading the 2009 LED Lab thread and many others while trying to catch up with KNNA's trail breaking. Luckily I enjoy it and LED research is something I prefer to do with my free time. It is one of those things that takes a big initial investment but then it pays you off with no more input.

Plenty of excellent grow journals on CannabisCafe.net. They are in Spanish but Google can do a fair job at translation and there are lots of great pics. This journal has a sizeable 1.5 gr/watt yield in soil:

Thanks Roor! I am checking out your Lowryder LED grow

Hey SupraSPL! how are things going? loved your lamps and your plants seem to like it too!

are you going to keep the updates of your grow here? hope so!

i tried to contact knna but never got a response, but i sure hope he is making a store for his lights!!! that would be excelent since i just can't build one of these panels myself.

good luck on your grow!!

deleted post