Let's talk DIY lights


Actually, for all the benefits of strips over COBs in terms of mounting, I see the Bridgelux VERO29s versus the Bridgelux EB strips have a distinct advantage in terms of lumens/$. The Gen2 strips run around 320 lumens per dollar, while the COBs run around 530 lumens per dollar. That is running both at the half-max test current. That’s a pretty big difference! I wonder why? Is the cost to make the strip that much higher?


@1BigFella I appreciate the input, and car terms makes perfects sense.

I’m still in the learning phase of this diy lighting thing and the question about the differing voltages was more of a clarification type thing. Just seemed odd to me I saw so many different Fv on LEDs in the same line.

Hell maybe it was separate lines. I’ve been up late for the last week or two reading this entire thread and browsing digi-key like my life depended on it.

That combined with generous amounts of Jack Herer and things start to all blend together.

I will say I’m leaning heavily toward the Vero29s esp since I already have heatsinks I’m pretty sure will work, a little more research is need to make sure. That offsets some of the initial costs getting them setup.


@Slackoff I run the Vero 29s and personally for me I won’t consider other options at this point. The light strength vs running less efficient doesn’t bother me because I have cheap electricity. I am insanely happy with my yields and bud density at this point and was worth my initial investment 10 fold.


@Daddy still so new I don’t know how to make a solid power consumption vs light output comparison. Probably something I need to consider, but, I have a 10Kw solar system and I get next to nothing from my power company for the surplus…so planning to put that extra ~2500Kwh to good use.

Do you have a journal or anything going you can link me into. I’d love to see and know ore about your setup.


In the case of Vero29s, they are the same line. Even maybe the same number of blue LEDs in the COBs. They are just wired up differently so circuit designers have three options. For example, I started with the high voltage COBs, but moved to the 50 volt ones so three in series would total 150 volts for my simple driver circuit. That uses no inductors and about $10 worth of parts for a 90% efficient driver.

You can look at my Drivers on the Cheap thread if you are into electronics stuff. Please be careful. If you have no electronics background, you should just go to HLG, Timber, RapidLED, etc, and buy some lights.


I know from reading several of your posts the stuff you mention is well above my skill level.

Regardless I’ll pop over and check it out. Cant hurt to get the exposure to some stuff. I promise I won’t attempt anything therein :shushing_face:


@Slackoff Daddy's test run, maybe a little something for everyone is my journal


@Daddy light looked familiar. Skimmed through first 60-80 posts then jumped to the end.

Wow. Nice haul!



RapidLED has some nice kits that you just plug together, so there’s little chance of blowing anything up or electrocuting yourself. If you use a Mean Well driver, the light wiring is just a simple series circuit just like the old Christmas light strings, except you have to get the polarity right.


@1BigFella Been looking more and more at piecing something together myself as opposed to a kit. I like the idea of saving some bucks and making something.

I’m sure I can fab my rails and wire things together. But when it comes to the finer points of electrical work I could for sure zap myself. So no making my own “simple” drivers. I’ll let someone smarter than me sort that part out.

Where I seem to really get hung up right now is sizing COBs to driver and really understanding some of the lighting efficiacy - efficiency stuff.

Do I just size amount of COBs based on listed Fv / into driver output?

Data sheets seems to show voltage rising as drive current increases?

If so do we need to account for that, or do we simply leave some headroom on the overall driver output to account for it and call it good?

Dumb it down for me, I’ve probably read or saw this explained but I’ve been trying to absorb so much and it’s largley foreign to me, it doesn’t all stick.


It’s pretty simple. You decide what color temperature COBs you like and how hard you want to drive them. If you look at the Vero29 datasheet (via Digikey) they have a table of example currents and how efficient the COB will be at that current. I like to drive them at half of maximum current or less because they get real efficient then in terms of lumens/watt. You want somewhere around 160 lumens/watt or higher. HPS runs around 133 lumens/watt, just as a reference.

So say you decide to drive a 4000 K XRC-40E10K0-C-7X at 1710 ma drive current. This will give you 161 lumens/watt and drop 69.4 volts. (These numbers are right out of the datasheet.) You will need a driver that can supply 1710 ma at 69.4 volts. Since it will be a constant current mode driver, it’s okay if it can supply more voltage as long as 69.4 is in it’s usable range. You use the pot to set the current and then the driver will increase or decrease the voltage to keep 1710 ma flowing. Voila: You have designed a light.

Of course, you probably can’t find a driver that can supply that low a voltage, so you string together two COBs with a total voltage of 138.8 volts or three with a total voltage of 208.2 volts and find a driver that can supply 1710 ma at one of those voltages. You have now designed a multiple COB light with a driver you can buy from Arrow or Mouser. You can find the capabilities of all Mean Well drivers by looking at their catalog or even on several websites devoted to them.

You can use more or less current: Raise the current and the forward voltage drop goes up a bit. (But not much.) Higher current puts out more light but at lower efficiency, so there’s a trade-off the designer has to make. Higher current also means more watts so you need more heat sink, but you can add a little fan to blow through the heat sink and increase its heat dissipation by 4 times!

Designing the heat sink may be the trickiest part of this. You want about 16 square inches of aluminum per watt unless you use a fan and then 4 square inches per watt is okay. Be sure to count both sides of each fin in calculating the total area. Vero29s put about half their power out as heat, so running one at 118.7 watts (again our example) means you have to dissipate about 60 watts with about 960 square inches of aluminum. That’s a 15.5 inch square plate, but we usually use something with a bunch of fins to make it a lot smaller. Or you can just buy RapidLEDs 140 mm pin heat sinks for $19 each.


@1BigFella You sir made it all seem so much more clear, cannot thank you enough!

I am pretty sure of the area of the data sheet you referred to. I will give that another look, with a little better understanding.


With all the diy lights somewhere on here a heatsink that was diy with aluminum nails was made. Drill holes insert nails solder. For anyone making their own driver or light it’s doable.


You could do that, but it’s important to use an aluminum base plate and aluminum nails. Most nails are steel, which is very bad at heat transfer. Another idea is to used nails with big heads like roofing nails and just glue the heads to the base plate with thermal transfer epoxy. I looked at that and even found some good nails and epoxy, and then decided the RapidLED pin heat sinks are a better use of my time. Partly because they come pre-tapped for mounting popular COBs and connecting to rails.

But by all means, give it a try and post results. It would be interesting to see.


a little too much diy to the extreme there. Heatsink all in all are pretty in-expensive. The cost could quickly add up buying a bunch of these specalty type passive big pin sinks

PC CPU Heatsink and fan combo is likely what I’ll what route I’ll go. A little extra wiring for the fans but it’s a pretty quick and simple fix. PC over clocking guy have all sort of indenpent 12v out with a wide range out ma out ranges as well.

Mine are all free but the same ones are easily had all day long on eBay for like $5-7 shipped in many cases


I have been lurking on the forum for a couple of years. My setup is two Mars 300 pulling a round 200 wats at the wall in a 2x2x6 foot custom box Hydro. After about 6 grows with this setup I am ready to upgrade. Going back to a post by @dbrn32 ( post 30 July 17) which was a 200 watt setup with EB strips

[quote=“dbrn32, post:30, topic:13186”]
"The eb strips are available in 3 lengths. We’ll call them 1 foot, 2 foot, and 4 foot. The actual measurements are 280mm, 560mm, and 1120mm. So we catch a break and can actually have room to mount a 4 foot strip in a 2x4 tent. If you’ve ever tried to jam a 4’ fluorescent fixture in a 2x4 tent, you can really appreciate this! Anyway, the voltage and current ratings of the strips vary with the length. For now we’ll focus on the 2’ version for our 2x2 tent.

The bridgelux data sheet says the fv is 22.1 v at test voltage of 700ma producing 156 lm/watt in 3500k. Let’s note that this is about 10 ish lm/watt higher than cxb 3590 we looked at. Max current is 1.4 amps and cost is about $7.40 per strip (digikey). So let’s see the cost of putting 200 watts in our 2x2.

The meanwell hlg-185-700 is a 700ma driver with max voltage output of 286v. Dividing the 286 by 22.1 volts of our strip, tells us we can run 12.9 strips. My gut tells me it would do the 13, but we’ll stick with 12 for now. Reversing the math, 22.1v x 700ma gives us about 15.5 watts per strip. Multiply the watts per strip by 12 strips, and we’re at 185 watts. Less than the cxb build, but let’s not forget that we had a little more lumens per watt. For now, we’ll call this set up comparable to the cxb and see if we can match up better with a different driver later.

As a big added bonus, bridgelux says we can run these without a heatsink. Weird right, I just spent almost $80 on heatsinks for my cxb… but the data sheet says they binned at 50c, so I’ll buy it. Buy we’re gonna give them a little help with the heat. So let’s try some 1.5"x 1/8 aluminium flat stock. (It will work I promise) but feel free to go wider, thicker, or opt for chanel or angle. The more surface area you provide the cooler it will be. Also, the strips come with poke in style connectors, so fancy holders are needed. Let’s look at pricing…"

12 x 560mm eb strips at $7.40 = $89
1 x hlg-185-700. $50
24’ 1.5x1/8 aluminium stock. $50
25 m roll of adhesive heat
Tape. $15
Misc hardware and wire. $20

Total. $225
By moving up to the hlg-185-1050 driver we can drive the eb strips a little harder. Still considerably less than the 1400ma max at a slight loss in lm per watt.

The 1050ma driver has a max voltage of 190. The increase in current will raise our fv of the strip a little. Off the top of my head, I think it’s about 23v. I feel confident enough in that number at this time. So doing the math, 23v x 1050ma= 24 watts x 8 strips= 192 watts. So we just increased our wattage using less strips and flat stock. Subtract 4 strips and 8 ft of aluminum. Congrats, you just created top notch led power for under $1 per watt.

My questions are: is this still a good upgrade or have things increased with new products and strips, Are there better choices like cobs, QB boards etc. If anyone has better ideas I am open to all

@dbrn32 @ktreez420 @Countryboyjvd1971 @bob31 @Myfriendis410 @MAXHeadRoom @Aolelon


I have the gen 1 EB strips in 3,000K in my flowering space. I love it; quiet, powerful, white light, adjustable etc. This year there were a couple of newer offerings but I like the strips for close, tight canopy coverage. I’ve had these down as low as 6" from my plants with no ill effects (other than phenotype issues).


The EB strips are a good bang for their buck, I highly recommend them and I know @dbrn32 likes them as well. Really it’s just what you can use for your space. Right now things are pretty competitive with strips, QB, and cobs just where you want to save some money.


If looking specifically at vero 29 gen7, most are available in all color temp options for each voltage class. Would simply be a matter of getting the right part number in each. I’m sure there’s a couple of exceptions, but finding specifics in stock should be only issue should you want to do something like that.

Also, something that may have been missed in the cost of any of the strips vs cobs is the other components required to complete build. Someone who has access to heatsinks maybe it’s a little different one way or the other. But if you’re buying everything, seems like strip builds are ending up cheaper in a lot of cases. And if you really want to look close at your build, make sure to account for any active cooling energy into your efficacy. If a data sheet lists a cob at 160 lumens per watt at 50 watts you figure your flux would be somewhere around 8000 lumens. If you need another 6 watts per cob in cooling your heatsinks your system efficacy drops to about 142 lumens per watt. Weighed against buying a bunch of passive pin sinks it may be a small price to pay. But sometimes you can also end up in the same place using a cheaper cob and the passive sinks and have less overall power consumption for a similar amount of light. If you have the time to do the research, it’s all worth looking at.


@dbrn32 sounds like I likely had digi-key up and filtered all the Blx line and didn’t do a sort by Fv.

I’ll give it another look with a fresh eye and a proper sorting.

I don’t have my mind made up one way or another so far. Still trying to get a good idea of what it takes to piece some components together to form a light setup.

For whatever reason the entire Blx line caught my eye, could be the SE COBs push in wiring combined with seeing a couple others builds. I also dig the idea of strips as they seem super easy to throw together and get going.

All I’ve been doing is research, so much so I had to take a couple days off, watched a few GM5 vids and restart this thread. All started to make more and more sense.

The efficacy is more or less lost on me, as you described. Hold on are you referring to the fans being 12v and generally around about 0.5 -0.7ma?

That would make sense, but I don’t see the point for concern. Perhaps I’m missing the obvious.

But that is a small amount on power to worry about, and they will all be on their own 12v supply.

Still so much to learn…