Lap Does LEDs - My DIY Quantum Board Cabinet

Brother Bear said:

Feckin schweet !
Can't wait to see em up and running.
BTW, this is the first thread I've been more interested in the lights over the plants !
Inspiration !!!

If you were to do a larger room, say 8' x 8'.
Same technique, just larger boards to cover the room ? Say four 4x4 ? Or 2 4x8 ?

Click to expand...

Glad to have piqued your interest bro! I am totally geeking out on these things.

Personally, I don't think I would go over 4x4 for the frame. Maybe a 4x8 if it was built right where it was going to be hung and 2 people were hanging it. It's aluminum so it wouldn't be terribly heavy, but that large can get kind of awkward I'd imagine. Not to mention I think the largest drivers available are 600 watts and that's what I'd use for a 4x4.

Dislexus said:

More blue light = leafier flowers/more trimming though? Will you be running the same clones to see any difference? What will give better flowering results, more overall lumens from the extra blue or a greater red/blue ratio at comparable lumens? hmm I can't wait to find out.

Click to expand...

I'm not real sure. But I seem to remember dudes a little while back adding MH to their HPS bloom rooms and they were getting a bunch more resin compared to without.

Dislexus said:

If you want the new build to be as efficient as possible you could oversize the wire as much as you judge practical. Are there quick-connects that can be soldered on the strip so larger wire can easily be used? Also if you're going to drill any holes to run the wire kinda hidden you can get some bushings or grommets to protect the wire from the edges of the hole and just make it look sexy too. I was looking at other builds and they weren't bothering but bushings/grommets are so cheap why not. And "cable tie mounting pads" to stick inside the c-channel to tiewrap the wire to.

Click to expand...

I'm using 16 gauge wire for the connections, to be honest, I'm not sure if that's just enough or overkill. I was just gonna see how warm they are after 12 hours of running them for the first time. I'm not an electrician and I rarely do soldering stuff so I was kind of nervous about doing this one. The contacts on the ends seemed really small and close together. I finished it all last night but I wasn't really comfortable today with the job I did. So I got some heat shrink and re-soldered the connections and I'm much happier and confident with the work. I tested with a multimeter and nothing seems to be touching the metal frame and I can get a reading from end to end on + and - sides.
So the heat shrink on the ends is about as pretty as this one's going to get, but I like your ideas.

Dislexus said:

I'm holding myself back from such purchases/builds until 2018 but damn its tempting. Is there a way to confirm diodes from China aren't counterfeit? Like does Samsung have a list of resellers, or something?

Click to expand...

Not really that I am are of, to both of those questions. Buying from alibaba is a lot like buying on ebay. Alibaba does a great job at giving you as much info about the vendor as possible. After what they give you, it's up to you to find one you trust. The vendor I went with has been selling on alibaba for 8 years, have had their facility inspected, they operate right in the Shenzhen district, and they do $1-2.5millionUS in revenue per year. My instincts tell me that they wouldn't be around for that long doing decent revenue like that if they didn't put out a decent product with components they say they are going to use. I also know of at least one other grower who has purchased from this seller. I am no expert, but as closely as I can see, these diodes look EXACTLY like the ones I bought from Digikey.

At this point, I'm not too too worried about how wiring looks, so long as it's safe. I'll just be happy if it turns on and I don't get shocked tomorrow when the driver comes .

Well I didn't get shocked and it did fire, eventually. When I planned everything out I fucked up my math somewhere along the process and I ordered the wrong driver. My fucked up math told me I could run all the strips in parallel. I plugged the driver in and they were nice and bright at the beginning of the chain, but got real dim at the end. It took me all day but I finally figured I had to break the 14 strips into 7 groups of 2 strips that are wired in parallel in each group...and find another driver. It turns out that the driver I'm using for the first rig fits perfectly with this setup and drives them really soft. It's crazy because it definitely seems much brighter than the first.

Man this is so much fun and these things are so fucking cool -

I can not wait to try some LEDs....

Yeah I remember people putting MH in their bloom rooms for the trichs, some were saying it was blues some were saying it was UVb? I forget. I see others using 18 AWG wire for similar projects so 16 AWG should be good.

My DC is rusty but I think you were originally wired in series hence the voltage drop? If all 14 ribbons were wired parallel then the voltage would fall across them equally, its when you get to further down the line in a series circuit that the voltage drops from resistance. When you split them into 7 groups of 2, the 7 groups are wired in parallel and the 2 ribbons in each group are wired in series? So the voltage is applied equally across the 7 groups in parallel? Maybe I'm confused but possibly your initial calculations were right in regards to 14 in parallel but you were wired up in series instead? I'm only going by one half of your wiring that I can see.

Looks like the quick-connects don't lay flat. The end diodes kinda humped up would just bug the shit out of me, plus they can only handle a limited amount of power/amps, screw that. Solder with heatshrink is the sexy way to go.

I love it and I can't wait to fuck around with it. I've done some recessed accent lighting with the LED tape-ribbons (looked a lot like yours) a couple years ago run in U-channel with some diffusing lenses over it, it WAS fun and came out looking so sweet. I wanna see those plants under these new lights. Using the old driver what are your calculations with the new lights/wiring?

Shit I was just poking around and saw CC-type LED ribbons (constant current) that don't have all those resistors so less heat and voltage drop and power consumption: https://www.comilighting.net/News/Product%20News/2015-11-17/7.html hmmmmm there's gotta be a catch

2 weeks or so left until harvest? What are you going to do with the 3500k strips you were using, put them in the house?

I am so fucking excited

Lapides said:

Crazy that there are still like 3.5 weeks left to go with the way these things are looking...more and more pleased every day..

View Image

View Image

View Image

Click to expand...

EPIC!!

I found for best math, I figure out how much forward voltage, Vf, in one segment, for instance in mys trips a segment holds seven diodes for a Vf of 20.3v I found that seven segments in series equaled about 143vF. enough to fill a 185H-C1400B driver. to drive the diodes at the desired current is where it gets tricky. By selecting seven parallel strings you will drive the diodes at 100% or 200mA per diode string x seven strings or 1400mA , too strong. To reduce the current you add an extra parallel string so that the current demand is more than the the available current. For instance in this current setup if I add an eight string the total current draw is now 1600mA. 1400/1600 = 87.5% real close to the target current of 150-165mA for maximum lm/w around 200lm/w real nice.

So eight parallel strings of seven segments or 49s8p is the described formula which will yield a very bright fixture to the plants and energy efficient too.

This is the same design approach I used for my GrowGreen boards designed around a C1050A/B driver with 184vF and 1050mA driving the diodes at 150mA each getting real close to that 200lm/W mark. I do the same with my strips too.

Yeah man that's good info. It's still a lot to wrap my head around but the more examples I do the easier it's starting to come.

I snapped a few quick pictures tonight. About 2 more weeks to go give or take a couple days. 8-10" stacked tight colas in a tiny cab. Couldn't be happier -

Lapides said:

Yeah man that's good info. It's still a lot to wrap my head around but the more examples I do the easier it's starting to come.

I snapped a few quick pictures tonight. About 2 more weeks to go give or take a couple days. 8-10" stacked tight colas in a tiny cab. Couldn't be happier -
View Image
View Image
View Image
View Image
View Image

Click to expand...

You are a pro!

AvidLerner said:

I found for best math, I figure out how much forward voltage, Vf, in one segment, for instance in mys trips a segment holds seven diodes for a Vf of 20.3v I found that seven segments in series equaled about 143vF. enough to fill a 185H-C1400B driver. to drive the diodes at the desired current is where it gets tricky. By selecting seven parallel strings you will drive the diodes at 100% or 200mA per diode string x seven strings or 1400mA , too strong. To reduce the current you add an extra parallel string so that the current demand is more than the the available current. For instance in this current setup if I add an eight string the total current draw is now 1600mA. 1400/1600 = 87.5% real close to the target current of 150-165mA for maximum lm/w around 200lm/w real nice.

Click to expand...

Never heard the forward voltage term before, by your description I guess it means the voltage drop from the resistance across the segment you're measuring? Even with short segments you'll still get some voltage drop so the diodes would be somewhat dimmer at the end?

Are the CC (constant current) ribbons just basically set up so that each diode is in parallel instead of series? Still trying to poke around on those.

Samsung released a new diode the lm301b that is 2.3% more efficient than the lm561c, but I can't find any that have CRI higher than 80. Some say CRI 90+ is better for horticulture?

Those colas look fantastic Lapides.

Dislexus said:

Never heard the forward voltage term before, by your description I guess it means the voltage drop from the resistance across the segment you're measuring? Even with short segments you'll still get some voltage drop so the diodes would be somewhat dimmer at the end?

Are the CC (constant current) ribbons just basically set up so that each diode is in parallel instead of series? Still trying to poke around on those.

Samsung released a new diode the lm301b that is 2.3% more efficient than the lm561c, but I can't find any that have CRI higher than 80. Some say CRI 90+ is better for horticulture?

Those colas look fantastic Lapides.

Click to expand...

Vf, Forward Voltage is the maximum voltage available to the circuit. Vf determines how many watts the driver can put out. A diode has a required voltage to drive it that is called the Forward Voltage in Constant Current systems. Volts add up in series and amps remain the same in series. Vf for the Samsung S6 bin is 2.9v. 49 diodes in series is 49 x 2.9 = 142.1v less than the Vf of a meanwell C1400A/B driver.
diodes on a strip are in series. Each segment is seven diodes each which has a Vf total of 2.9 x 7 = 20.3v total for a segment and seven segments is 142.1v close to the max on the driver meaning it will drive close to 200w the max output of the driver.

each series strip of seven segments can be connected in parallel. Amps add up in parallel. So if I want to drive the diodes at 200mA each I would set up enough parallel strings to add up to 1400mA or seven rows. But I want to drive the diodes softer than 200mA to get more photons out of them closer to 200lm/w is the max at about 160mA. By adding an extra row I now have a demand for 1600mA but the driver only delivers 1400mA so less than 200mA per row will be delivered 85% of 200mA or 150mA just where we want to be.

So in summary seven segments in series is 49 diodes with 142 Vf. Eight rows of 49 diodes is 1600mA with a total of 343 diodes running at 150mA about 200lm/w x 200w = 40,000 lumens at 200w quite impressive.

Lapides said:

Yeah man that's good info. It's still a lot to wrap my head around but the more examples I do the easier it's starting to come.

I snapped a few quick pictures tonight. About 2 more weeks to go give or take a couple days. 8-10" stacked tight colas in a tiny cab. Couldn't be happier -
View Image
View Image
View Image
View Image
View Image

Click to expand...

Those plants are amazing. great grow. I can almost smell them from here. you have some great grow skills. If you want to try a few of my 200w boards they run on a 1050A/B driver and cover a 2x2 area nicely. The GrowGreen boards are very inexpensive for what they do. I don't sell them per se, it is a co-op shared cost which keeps costs down. A 200w board is $60 plus $20 S&H multiple boards same shipping cost.
I am doing like you. I have have strips and wanted something more powerful so I designed a board that works just like I described earlier.

Just a great grow, your skills are def in growing
namaste

This is awesome and very inspiring.

Your plants look really good, well done

You can get 3m double sided tape and do it yourself. I use 1/8 th aluminum and no heat

I took a couple branches down on sunday and looked at the buds under a usb microscope. I was really surprised to find like 90% cloudy trics, about 7% clear and 3% ambers. I'm only around day 60 or so but fucking went ahead and chopped today. I usually let the glue go 68-70 days. I checked my calendar 4 times and was wondering if maybe I was looking for an excuse to harvest early since I haven't had any of my own weed in ages. But I asked a buddy who grows under leds who has grown under HIDs in the past and he confirmed his plants also finish earlier under leds.
Just another amazing benefit from these things. And it's only gonna get better and cheaper. I will never buy a hid lamp again.


One glue plant -








A few of the nicest colas -


I forgot to add. I got a Kill-A-Watt thing to see what I'm pulling at the wall and it was 320 watts fully cranked, which it was the whole time. I should do alright from the looks of it. I think I said in the beginning I'd be happy to get a couple ounces to pay for the experiment. I think I did.

AvidLerner said:

You can get 3m double sided tape and do it yourself. I use 1/8 th aluminum and no heat

Click to expand...

I've been trying to figure out what you mean. Can you elaborate?

Laps did you finish it with the original 3500k lights or switch out for the 4000k? I'm guessing went all the way with the 3500k because that 4000k rig looked wider than 1.5 ft?

Looking forward to the total dry grams and gr/watt grand finale. Gr/watt will probably skew down due to the stress test you were running but that just makes me more curious.

late but locked in

Lapides said:

I've been trying to figure out what you mean. Can you elaborate?

Click to expand...

the suppliers use double sided 3m tape to adhere the strips to anything. you can do the same.

nice harvest. I find the time changes with leds vs. hps. faster to finish

Lapides said:

But I asked a buddy who grows under leds who has grown under HIDs in the past and he confirmed his plants also finish earlier under leds.

Click to expand...

AvidLerner said:

I find the time changes with leds vs. hps. faster to finish

Click to expand...

I wonder if HPS puts out 730 nm far red, and if so, is the reason for this difference... or is it the relative extra blue of even the warm LEDs... or the 730nm of the LEDs just being overwhelmed by everything else its putting out for the plant...

Having a week give or take shaved off of flowering times without having to do a damn thing is kick ass.