Let's talk DIY lights


I think I’m loving the strip idea. It seems like my best option for light coverage, and light mobility. One day soon I’ll move into a bigger place, and I like the idea of multiple fixtures so I can move them around and put them where I want them to be.

I just have to decide and make a decision lol! It’s hard @dbrn32!



If you want to do something with the strips, the biggest decision is how you want to cover your tent. You can pretty easily get more or less output from the sane shape of fixture. For your veg tent, do you wanna end up with 2 fixtures that are roughly 2’x4’, 4 fixtures that are 1’x4’ or 2’x2’? If you’re not sure about the future, you can always reconfigure them later. What works best right now? Once you decide on a size, figure out about how much power you want there and we’ll make it happen.

If the quantam boards and heatsinks become available, that can become a very very easy build.


@dbrn32 another thing I phrased very poorly in my comment was, I’m aiming for 600ppfd/sqft. Is that what you figure or does that change your response XD sorry I wasn’t clear on that. I’m at work right now and might watch some more videos on the subject during lunch, so I could always do the math myself


I’m pretty sure all of those lights will provide an average of 600 ųmols or more. You may have to run a strip or quantam board build a little closer than the cobs, but they should all make it.


Or maybe I’m confused on how I saw ppfd of a hps worked out. Idk I’d say disregard that last comment for now and I’ll refresh what I had in mind last night in bit. Lot of info to take in XD


Search quantam board ppfd on youtube. There is a good video there that shows them testing a few variations


@Carnako I’m speaking your terms right? Ppfd=ųmols per m/2/s?

I’m not exactly sure what the hps numbers are that you’re trying to replicate though. Basically you’re trying to target a certain amount of photosynthetic active radiation over your canopy, correct? It’s a much better method than saying watts per foot, because it considers the wavelengths of light that plants actually use. One of the benefits of led’s is that we can use multiple lower powered sources to keep this measurement at a fairly even amount across the canopy. Targeting 600 ųmols is pretty solid. Considering the actual amount directly under any light source will be higher than the edges, you will always get some help from reflection in a tent. This being said, there’s a limiting factor that keeps people from throwing 1000 watts of cobs in a 2x4 tent, it’s co2. Probably heat at that point as well, but we don’t really need to get into that.

I believe the general consensus is to target somewhere between 600-1000 ųmols. More or less will still grow. But you will be out of the window of maximizing your energy/yeild ratio.

I’m pretty sure all of the builds I mentioned would give you at least the 600 ųmol average, probably a little higher like in the 700-800 ųmols average for most. But you said at least 600 right? But still well within the window. I would recommend a meanwell hlg series a or b with any of them, will provide you with the ability to dim the fixture anytime you needed to, or when you wanted to run a little less power.

I think as far as cost, the cree build would probably be the most expensive. Depending on driver configurations, the citizen and quantam board or a samsung strip build would probably be somewhat close. And the vero 29 and eb strip we can usually get around to around $1/watt would probably come in the cheapest.

The cree, citizen, and quantam boards would probably come in as the most efficient in this scenario. So you kinda pay for that premium, but I also think that you’re getting a little more even coverage and some extra power to boot. So take that for what it is.

There are a few that offer kits that should be able to replicate any of the cob builds. You pay a little more than sourcing yourself, but it’s hassle free and you still save some $ assembling yourself.

Like I said earlier, the quantam boards are a really easy option here too. But last I knew the boards and heatsinks you would want were out of stock. I think they can/will tell you when they are expecting more, and even put you on a list. If that’s what you decide.

All things considered, it’s pretty easy to give you the current options. But probably best that you plenty of snooping around the Web and make sure you get what you want.


Ok Yea that cleared up my confusion, just kinda flustered this morning worrying about getting this big truck unloaded and health insurance. When I get home I’ll take a more detailed look at the stuff you’ve mentioned. Is there any difference or research done with like gram/watt in these different LEDs? Or are all they the same because they are putting out similar quality of light?


Yeild info is out there in different grow journals around the web. You kind of have to take them with a grain of salt though. I would consider yeild potential about the same amongst the type of lights. You’ll maybe yeild more or less overall depending on current level chosen and how much coverage of each light you have. But it’s definitely not like one of them will put you at 2 grams per watt while another would be limited to 1 gram per watt.

I think the experienced growers are pulling around 1.5g per watt once they figure out their setups. Strain will have a little to do with that too. If you’ve watched some of the growmau5 videos, he’s knocked on the door of 2gpw with the 3590’s and quantam boards. So the potential is definitely there.


I love the idea of 2 - 2x4’ fixtures! I’ll never have a tent smaller than 4’ so that’s ideal for my setup.



Ha, that address @ktreez420!

Like I was figuring for @Carnako, 10 44" at 700ma or 8 44" at 1050ma. Or double each for 22" strips and you can place them the opposite way.

You can run more or less, but either of those should put you in the wheel house for optimal par range. I think 8 at 700ma would be ok, but you’re literally talking about and extra $30-40 to add another 60 watts at that point. The 700 watt setup is going to run a bit cooler and provide more light per watt. So the choose is there to save a couple bucks, or go with cooler more efficient light.


I’m all about the coolest (temp wise lol), most efficient lights. That’s what I’m going for, and good light coverage obviously. I love that I can put the strips closer than COBs. I’m a little confused though on what you were saying about this;



No problem amigo.

The 8 would be amount of strips, 700ma would be drive current. Basically, the lower drive current that you give to your led’s the more efficient your led’s will be. Max current on the 44" inch strips is 1400ma (1.4 amps) that suggests that per manufacturer data, each strip could be run at 62 watts. At this point you would have the most available output from the strip. But you would also have the least amount of light per watt, the most heat, and shortest lifespan.

Stepping down from there in the meanwell driver constant current line, will put you at a drive current of 1050ma (1.05 amps). From here, you will lower the actual wattage per strip to around 47 watts. But increase the amount of light per watt and lifespan, and lower the amount of heat. You can find some people running builds at this current, and they seem fairly happy.

The next step down in the driver line is 700ma (.7amps). This will run each strip at about 31 watts. Again, another jump in lights per watt and lifespan, even less heat. This also happens to be the current that bridgelux bins them at. Their data sheets tell us that this is the test current for binning. The sheets aren’t always exactly the easiest to decipher. But it looks like they binned at 159 lumens per watt here with a tc temp of 50 degrees C. I’m guessing that they were trying to support their claim of not needing a heatsink. The tc temps I measured were about 38 C when mounted on 1.5"x.125 aluminum bar with the double sided thermal transfer tape. Less heat would mean the actual light produced would be more. Probably not much, but any increase is better than a decrease. They also rate the lifespan at greater than 50,000 hours at this point. You can expect some drop in performance over time, but again the lower current will equate to higher performance.

You could continue to drop the drive current and see efficiency numbers get better, but they’re also not exactly linear. So we’ll assume that since the test info bridgelux provided was at 700ma, that’s the “happy medium” for cost vs efficiency. It wouldn’t be unrealistic to sew a number close to 180 lumens per watt when driven at 350ma, but you’d also only be producing about 16 watts a strip, and would require a whole lot more strips to get par levels up. So for our purposes the 700ma and 1050ma are probably the best options.

So the comment about 10 strips over that area but 8 would do, referred to how many strips over that area to hit the target par levels that @Carnako was looking for, and probably a good target. Last night I dug through the meanwell driver catalog and looked at a few different configurations, there’s not a perfect match for 10 strips. The connectors on the strips are only rated for 300 volts, so that really takes all of the bigger drivers out of the conversation. There are options that run 1 strip, 2 strips, 4 strips, or 6 strips.

If you’re shooting for 10 strips, 2 of the 200 watt drivers that push 5 each, but they’ll be more efficient running 6. It can be done with a 200 watt driver running 6, and a 150 watt driver running 4. The price difference between buying 2 200 watt drivers and 1 200 watt and 1 150 watt is literally $3.

The math works out like this

31 watts per strip at 10 strips is 310 watts over your 2x4 fixture.

By adding the $3 and 2 additional strips the build cost goes up about $30, but your 12 strips will run about 372 watts over the same area. Provided more par, better canopy coverage, and because the driver would be nearly full an increase in driver efficiency.

The two 200 watt drivers running 12 strips at 700ma is what makes sense for getting the most out of your money, coverage, and efficiency.

If that’s within your budget, then the way to go with the strips. You will probably see an increase in price over the vero 29 build we were looking at. But you’re also going to get more watts and better coverage with about the same efficacy and probably a more electrically efficient package. That’s pretty much life in building for efficiency and coverage vs cost per watt.

Off the top of my head, you can still get the drivers from rapid led and take advantage of the discount. Should put them in mid $40’s each. The strips are $14.60 each so about $175 for 12 of them. Then whatever deals you can find on aluminum to mount them and build some sort of frame, thermal transfer tape, hardware, and wiring.


That’s the plan then, I’m going to start buying the pieces next week when I get my next paycheck. Just took care of rent, car, insurance, etc.

Thanks once again @dbrn32! So if I get the thermal tape that’s double sided, will I need to drill anything in (other than my frame obviously)?



I have a small cabinet (1.5’ x 2.5’ x 4’). I bought a 4000K 135 watt quantum board kit from HLG. I had to dial the light down because at max power, it was causing severe nutrient deficiencies in my plants. This, in turn, stunted growth during the flowering stretch. Once I got that figured out, though, I am extremely satisfied with it. It’s definitely a great light for my needs.


Next grow should be even better with it.


You got it @ktreez420! The thermal tape makes mounting the strips a breeze. I sourced my aluminum locally in 10’ lengths. So after I did some measuring and cut to desired lengths. I did some light sanding on the aluminum. Just the side I was mounting strips to. Aluminum stars to oxidize quickly, so you just want a clean flat surface to help the thermal transfer. Make sure you do this in a well ventilated area or wear at least a dust mask. Aluminum dust isn’t good to breath. Then a a quick pass with 240 grit followed by another with 600 grit. After that I gave each strip a good cleaning with some rubbing alcohol. I applied tape to the strip first, attempting to keep the 20mm strip of tape as centered as possible on the 24mm led strip. An extra set of hands could be helpful, but I was able to do it myself. Once the tape was applied, I ran my across it a couple times making sure I worked out the any creases or bubbles. Once satisfied, I peeled the protective film from the other side of the strip and applied the strip to my aluminum bar. Ran my finger down the edges of strip and that was it.

2 things that seem to really jump out at me as helpful… when you build your frame, you can rivet if you’d like, but use at least a few bolts that are a little longer than necessary and make sure that they are oriented that the extra length protrudes the frame on the same side as your strips. This will allow you to set the fixture down without damaging your led’s. The other, you probably want to mark your aluminum where you want your strips to be. Assuming you want them centered, if you miss, I’m not sure you could peel the tape very easily. I didn’t try, but I wouldn’t want to either.

The tape seemed to hold plenty good, especially after I powered light up. I don’t think I could remove them without damaging the strips anyway.


Looks great @mcpd_refugee!

What was your experience with the kit? I think the average guy could throw one together pretty easily, but I’m sure people get nervous about spending $ and not being successful.

Would you be willing to post some pics of your light for the members to get an idea of what it entails?


I don’t have any photos of the assembly process, but it was easy. I had everything put together in less than fifteen minutes.

I cleaned the heat sink with 90% rubbing alcohol and a soft cloth, then screwed the quantum board to the heat sink. Interesting to note that a thermal backing material or thermal paste is not required. I screwed the driver to the heat sink. I connected the driver wires to the included wires with the included wago connectors and then connected those wire to the connectors on the board. Everything is just using levers and pushing wires, so there’s no soldering. I then used the screw-in connector to connect the power cord to the driver. And that was it. I hung it and it was ready to go.

I think the total cost of the light was $178, so it was one of the cheaper options for me. I probably could have bought COBs for less, but I liked this option because it is very low profile and space is at a premium for me in my cabinet.



P.S. One more note: I am no electrician. I have wired ceiling fans and lights, and that’s it. So, if I can put a QB together, anyone can.


Looks great!

Personally I don’t feel you would’ve saved much if any by going with a cob option. Given the size of your space, the qb is probably a lot more versatile.