Let's talk DIY lights


I’m actually looking at 2 per tube for 5 total which ought to give me good coverage. I also figured that there is a LOT of surface area and aluminum has such a high thermal conductivity that I should be fine.


I tried to pick some up today but there’s a Delta IV launch today and he’s sitting on the console. I think it’s scheduled for 1:00 PM today.


I agree you’ll be plenty good, even if there were more strips. Looking at it from a coverage standpoint, I may even consider doing four bars of aluminum and putting three strips on the two outer bars and two strips on the two center bars. The middle will have highest density with help from the diodes on each side but you’ll have more diodes toward the edge creating a bigger overall footprint and more even overall light canopy.

You can do it either way and I think you’ll be happy. But the spacing I suggested will get you a bigger window to play with heights and also much less need to worry about a center hot spot.


Excellent point and it will reduce the weight by 20% too.


So WW doesn’t like the cobs any closer then 28 inches…


My Blue Dream has just grown up to within 11 inches of my Vero29 COBs. They are the high voltage ones running at about 1.22 amps, so they are about 82 watts each. About 15000 lumens. Plant looks fine.


Maybe they just need to be hardened to the different light source a little?


They have been at :100: for 2 weeks now both WW grew into the light a little bit and :taco: lol nice little curl on the edges too. I already seen that they can be a little finicky with heat and light so no big deal. Stepped the light up 8 inches and they will be fine. The other strains love the intensity.


My 30 watt photo red 660 nm COBs finally came in from China today. It looks like they each have 30 one watt LEDs on each one. They have about a 1.5" x 2" metal mounting plate on the back surface, so good thermal transfer should be no problem.

I am supposed to get a new camera arriving tomorrow. I’ll post some pics when I get one running.


Built a test circuit to drive one of these Chinese red COBs. It is supposed to run at 1 amp with 22.0 volts. Here I’m running it with 0.92 amps at 23.5 volts. So their specs are not exactly accurate, but let’s assume these LEDs can take an amp. My measured power is actually 21.16 watts. Not so bad for a “30 watt” COB. You really can’t look at it. It’s very bright looking straight into it. I had to move way back with the camera though it’s pointing straight up…


Sweet! Do you have a way of using your filters to test that its indeed 660nm? Not that we have any reason to doubt it, but I’m always skeptical of stuff like that without a quality reputation behind it.


Here’s my test board. 24 volt transformer, bridge rectifier, big honking capacitor, LM317, and a couple of resistors to give me around 1.25 ohms = 1 amp constant current. In theory, anyway. It’s actually 920 ma. In this photo I’ve just unplugged it and the LEDs are lit a bit because they are discharging the cap.

I checked the output with my red gel and lux meter. From one foot, the red COB reads 1200 lux. Measuring a Vero29 5000 K COB with my red gel from one foot, it reads about 9000 lux. If you look at the Bridgelux Vero29 datasheet, the spectrum graph shows the 5000 K has about 17% of it’s energy at 660 nm. One of these red COBs per White COB should bump the 660 nm energy up by about 1200/9000 = 13.3%. That’s not 13.3 + 17. It’s 1.13 x 17 = 19.3%

I do have six of them, but I’m running six Vero29s right now, so they are just going to give a slight improvement to the spectrum. Not as much as using 4000 K Bridgelux COBs, for example. Of course it’s only 21 watts, while the Vero29s are running at about 82 watts each. But 1/4 the power of a Vero29 for such a slight improvement sounds like a no go to me.


I have a little <$10 spectroscope I got off amazon. I think it’s just a diffraction grating (probably a piece of a DVD) and a scale but it claims to be accurate to 5 nm. Problem is I have nothing to compare it to. I need to get a real 660 nm LED. Then I might be able to tell.

I’ll probably get some XP-E HE’s from RapidLED for a comparison. I know they’ll be the right wavelength.


I’d send you a couple in the name of science if there was way lol. One of these days perhaps.


No need. RapidLED is really, really fast. I ordered some COBs a few days ago and they’re already here. And I will probably order a bunch to see if I can get my grow room setup for flowering.

I tried my little spectroscope and it doesn’t work with the red light. Maybe the peak is too narrow to see it or something. When I point it a fluorescent light I can see multiple color bands.


I got the 660 nm XP-E HE LEDs from RapidLED. Checked them against the red COBs I got from China. Both my eyes and my wife’s eyes say the China COBs are more orange-red compared to the Crees. I got my spectroscope to work and it says the Chinese COBs are 640 nm, not 660 like they said.

So I would not trust them to send the exact product you order. Doesn’t matter for me: I need some 640 nm LEDs as well as 660, so I’m fine. Just a word to the wise. Buy from reputable sources. If something seems to be too good to be true, it probably is.


Thanks for checking that! I’m always a little hesitant when buying unproven leds. You never know when you’re going to find a gem or just get something cheap.


Just starting reading from the start, looks like I have a long way to catch up. Lol
Great topic


it’s a long thread lol! There’s some really good information and some babbling, so feel free to ask any questions. There are some random projects flowing around ranging from probably $100-$1500 plus, so don’t feel like it’s a big deal or beyond small or beginner stuff either.


I think I’m liking my setup so far. Props to @dbrn32 with the help verifying my setup. 6 weeks and 29-21 inches.