Cree DIY COB LED? Why do people use white lights?

Wow, this stuff makes my head hurt (worse) and my eyes glaze over. We could really use a Homer Simpson level DIY COB primer thread (or... section, lol).



Are you talking about the (Citizen) 48 or 52?

I have gotten the impression through semi-random web searching that the efficiency of various COBs varies depending on the amount of power that they are driven at... IOW, that one brand/model might be more efficient than another at a certain power level, but when the power is increased, that the efficiency ranking could change. Is that correct, or not? If so... HtH do people ever figure out exactly what they should be building/running, lofl? I'm lost after the "Okay, my grow room area is..." step :rolleyes3 . At least with HPS, assuming a decent reflector, one can look at their area - and the shape of that area - and then generally determine what wattage and how many HPS lights it'd require to give good coverage and intensity for a room full of sativas (or indicas, I guess). Speaking in generalities, they can even get a rough estimate of n watts per square foot. Sure, some HPS bulbs (in a particular wattage) are better than others, but there does not seem to be a great deal of difference. With LEDs, though... Several manufacturers, each seems to have several models of COB - and I don't even know how to "read" the codes for one brand (FFS!), like Cree has CXA2520 all the way up to (if not more, by now) CXB3690 or maybe it's CXC or CXD I can't even remember that much... - which have different levels of efficiency, which again seems to vary by the amount of power they're being fed, and bins and drivers and stuff which is like... IDK. I figured I'd do a little searching and try to give myself a basic understanding and it was like I set out to figure out which brand of tv dinner was more nutritious and only end up finding articles on the genetics of what the freakin' mother of the cow that eventually went on to form (some small part of) the salisbury steak ate along with a bunch of electron microscope images of the plastic tray. I should have just bent over and ran full tilt into the nearest brick wall, lol - my head wouldn't have hurt any worse, I would have gained just as much of a basic understanding of all this sh!t, and I could have saved a few hours. MAN, I wish I was still twelve years old and hadn't gotten hit in the head with everything from baseballs, the ground, and a few 3,000+ pound vehicles that wanted to see if they could fit through my windshield...

I started out thinking that Cree was what you bought if you wanted the best. I don't even know if that's right. I saw mention of Bridgelux, Citizen, Osram, Philips, Vero, Nichia, Toyota(???), and even Sharp who last I heard couldn't even make a halfway decent microwave. IDK. Sorry for the rant, but... I guess you guys haven't been knocked out half a dozen times and knocked silly LOTS more and are still young enough that your cognitive function is still kickin', lol. But there's a lot of decrepit old farts here whose hands might still be steady enough to use a soldering iron who could benefit from a thread like "For a 3' x 8' Room, buy THESE Parts from THESE People for THIS Much Money for Great Intensity @ THIS Many Watts (THIS Year)." Then, people could advance from the pre-school thread to a kindergarten one that discussed options like what if your room is smaller/larger, or hey I cannot support that many watts in my house trailer so if I shrink my space can I get by with n watts and, if so, which changes do I need to make to the basic design - with part # recommendations, please? Stuff like that.

I have an Amare Technologies SE350+UVB here, that I've got to do a grow with. I'm assuming that I'll have to give it back because it's, what, something like a $900 piece of kit, lol. But I figured after the grow, I'll have learned something about growing under LEDs (as I have never done so before and my current knowledge level is... well, you just read my rant, lol, so that much is obvious ;) ) and ought to be able to do better and learn more on a second grow (and so on) . PLUS, with a little bit of experience, I might wonder if changing to a different style / training method might give a higher g/w . Stuff like that - IOW, I can see the one grow turning into many that could still be learning experiences for me and others... and potentially help the company's sales by showing that even a dumb@ss (in regards to this LED stuff, at least) like me who has grown before with different types of lighting but never LEDs really could make the transition to LED. Which means... well... when I complete that first grow, instead of emailing to ask, "Do you want me to ship this LED panel back now?" I'll probably just, err, wait on them to email me and in the meantime keep on using it and learning with it, lol. Now, of course, my possession of the panel after that becomes completely hypothetical - and I know this. But... IF I still have the thing, say, six or ten grows from now, I'll probably have reached the point of diminishing returns, that any further knowledge I get from it would be relatively small. So I was thinking about that hypothetical future point and wondering, "Okay, this panel has some kind of COBs in it. I think they might be (CXA???)2520s. Maybe 2530s, but I read that the early versions of this model had 2520s and this particular one was so early it might have been a prototype (or even THE prototype, lol). So... Could I upgrade the COBs in this panel at some point? What, exactly, would that involve? JUST changing the COBs, or would I need to change bins/drivers/power supplies/whatever else is in the thing I sure am not going to open it to find out because it's not broken and I don't want to change that, lol. I'm not even sure about this much, but I think that all COBs are not the same size, so I'd at the least need to figure out which one(s) is/are the same physical size.

Obviously, like I mentioned, that thought about upgrading is all hypothetical. But it sure would be nice if I could find a stickied thread that had the basic information - and in a dumbed-down form - that I'd need to digest in order to even think about it. Err... I'm ranting again, aren't I? Sorry.



That's the way that this SE350+UVB is: Six COBs and each one has a ring of ten mono-color LEDs around it. The COBs give "white" (I think it might be ever so slightly off-white, but my color perception is all screwed up and isn't even the same from one eye to the other, so I'm not sure and couldn't begin to say whether it leans a little towards yellow, green, blue, red, or... But I'm pretty sure that the individual LEDs are a mix of red, blue, white, and "IR." (And why everyone calls them IR is beyond me since last I heard the human eye cannot even perceive infrared but we can see the illumination from these, albeit dimly.) I think it's a really cool product, to be honest. Seems to have been well built, from some pictures I've seen it is capable of producing (and producing a lot for just ~350 watts), comes with removable 90° lenses for the COBs that really look like decent optics instead of cheap junk (and you get a 120°, IIRC, pattern without the lenses - which gives you some ability to tailor the light to the space/setup in terms of penetration vs. footprint size - and since there's one for each COB I suppose you could remove the two middle lenses and keep the four at the corners installed, or vice versa for a "hybrid affect"), and it was designed with the purpose of growing cannabis in mind. I couldn't say how it would rank in terms of other brands (and this model is a year or two old and a generation behind the company's current version in terms of wattage, COBs, etc.) but I'd think it'd fall somewhere nearer the top than the bottom at least. I'm pretty impressed already with just the impressions I have gotten with it before actually growing with it, feel privileged to have been given the chance to use it, and am a bit embarrassed that I was not able to get an a/c so I could have gone ahead and used it from the time that I received it instead of having to wait until the weather cooperated.

Now that I recall, the concept of adding mono-color LEDs to customize the spectral output (and, of course, the overall level of illumination) is not new. Six years or so ago there was someone who was either selling or getting ready to sell induction fluorescent grow lights, and he was talking about adding a few (like four or five on each end of the fixture, something like that) for that purpose. IDK what ever became of that, because he seemed to only have been here a few months and then quietly disappeared.

This post is quite a train wreck. Apologies....
Watch growmau5 on YouTube.

After tons of research I am building mine using citizen 1825s.

I will share my diy with you soon.

This forum doesn't allow sharing links so makes it near Impossible to help source parts like we doing other places.

Sad as it holds the community back based on $$$ and nothing else.
 
Have you run across Dion on another forum? There's a thread with international users and links to various distributors. Search for dion led diy.

And have you seen what Citizen has been doing? Great prices and performance, and I think they have UK distributors - I think Budmaster uses the Citizen COBs? :hmmmm: Their latest series nearly matches the CXB3590 for much less cost.
I will be testing the 1825s very soon.

I will be pushing 4@ 113 watts Max dimmable to 10% in a 3x3.

Very excited. The heat sinks I am using are just badass looking!
 
I had not really talked about efficiency in my post however that is one of the driving factors for not wanting to spend the much higher cost for the 3590s because I in in the works of designing my own light in CAD currently which will utilize much more efficient drivers and a better overall layout for use with COBs. On another note I am looking at something that seems like for the minimal investment would greatly increase the power of the stock light. That is a CXB2530 3000k 90CRI, 18 of those would fill out the panel quite well.
Citizen 1818 will save you ton of money.
 
I have done a ton of reading.

I just ordered 4x CLU058-1825 3500K and 2x drivers.

These will be serious beasts and will be pushed very hard on large heatsink.

Google Titan DIY Cob if you want to see my inspiration.

Right now based on sphere testing it seems citizen is the best bang for your buck.

Cree and Vero are no longer ahead of the pack unless your talking price.

Granted someone that hasn't read the recent tests will say I am wrong and most likely own cree cobs.

In your price range I recommend the CLU048-1818 and the spectrum is your call. Citizen jas no bins as they only sell the best COBs they can make.

Citi and Bridge lux have new models coming out soon.

Do your research.

Search Growmau5 and watch the tutorials then watchthem again.

HAHA, man! I was just looking at it right before I read this post and I am decided :D Thank you!
 
Hello, could I ask if you are purchasing the Citizen chips in Europe. If so could you pm me the place please? I find these at £55+ each over here which is more expensive than I pay for the Cree 3950, although I see them as cheaper than the Cree in the US. Thanks in advance :thumb:

Tomula good luck with your build, I'm sure you will be really happy to get back under white light instead of blurple :)
 
Hello, could I ask if you are purchasing the Citizen chips in Europe. If so could you pm me the place please? I find these at £55+ each over here which is more expensive than I pay for the Cree 3950, although I see them as cheaper than the Cree in the US. Thanks in advance :thumb:

Tomula good luck with your build, I'm sure you will be really happy to get back under white light instead of blurple :)

I've found that Octopart is the easiest way to keep up on who has what, and how much. :slide: Fire it up, enter your part number or search by maker, and up pops all the international vendors that carry it, along with the current inventory and price.

Currently, it looks like only CDI and Digikey have the CLU-048s. :straightface:

:Namaste:
 
Seems like the only way to get them in here is 47$ for the chip itself the 1825 3000k 90CRI. And postage is 48$ so if I wanted to buy 3 of them which is enough for 400W for me, that means 200$. Or I can get 2. And power supply for 150$. So I can get to 400 dollars with the heatsinks and a fans. But 200W per chip, that's some mean heat, but then the 3 pounds of aluminium and pc fan will do the job right? I still have a month or more before the build, but I'm trying to get all the info I can. As for the blurple light I have right now, it works, but the girls are stretching more than with the Mars. I think it is because of lack of the blue light, right? Also I have found decent deal on weaker Citizen LEDs with free shipping but my guts tell me that the stronger light will get better intensity. I dunno, this is uncharted territory for me. I like all the feedback from you. Thank you.
 
Seems like the only way to get them in here is 47$ for the chip itself the 1825 3000k 90CRI. And postage is 48$ so if I wanted to buy 3 of them which is enough for 400W for me, that means 200$. Or I can get 2. And power supply for 150$. So I can get to 400 dollars with the heatsinks and a fans. But 200W per chip, that's some mean heat, but then the 3 pounds of aluminium and pc fan will do the job right? I still have a month or more before the build, but I'm trying to get all the info I can. As for the blurple light I have right now, it works, but the girls are stretching more than with the Mars. I think it is because of lack of the blue light, right? Also I have found decent deal on weaker Citizen LEDs with free shipping but my guts tell me that the stronger light will get better intensity. I dunno, this is uncharted territory for me. I like all the feedback from you. Thank you.

Have you checked to see what the shipping cost is for Timber kits? At the price you're thinking of paying, they might be a choice.

Timber Grow Lights
 
Have you checked to see what the shipping cost is for Timber kits? At the price you're thinking of paying, they might be a choice.

Timber Grow Lights

Thanks for suggestion. This is interesting but I want one driver per one LED. And with 3 Citizen LEDs and 3x 150w drivers I can get into that price range and have 400W of light. That is enough for half of the tent I will be using. I guess 450$ dollars it is. And the photos under those lights will look awesome!
 
Thanks for suggestion. This is interesting but I want one driver per one LED. And with 3 Citizen LEDs and 3x 150w drivers I can get into that price range and have 400W of light. That is enough for half of the tent I will be using. I guess 450$ dollars it is. And the photos under those lights will look awesome!
Never heard of such a thing.

Your gonna need one large heatsink.
 
Well I guess I am pushing 2 cobs per driver at 113 watts a piece.

Main reason I went with 4 was for even coverage of the tent.

Also if you have 3x drivers for 3x cobs it will be difficult to dim equally if you plan on dimming.

The reason for 3 drivers is that if the driver goes on holiday other lamps will be still shining. And I plan on dimming. I will measure the power output and I guess the difference in light will be small. I am also thinking about light mover, so the coverage will be better with that right? As for the heatsink I was looking at 3 pound aluminium heatsink 4x4x4 inches big, 3 pounds. I still dunno if there is difference between 8x50 light, 4x100 light or 3x133 light. I know the spread is a factor, but 150Wx150 lumens is around 20000 and that's some serious light coming from the chips. Or I can get Mars Pro 2 for similar money. Still, I have time to think it through and thank you very much for all suggestions! :love: :Namaste:

Chladi_ALU.gif
 
Good thread.
After playing around with some various "retail" solutions to lighting, I decided to get building. When I add up the cost of good quality components, and compare to my platform size, I am saving roughly 60-80% overall from "full spectrum", and 30-40% from basic r/b combo setups. While the r/b do work, my application is vertical farming with leafy greens almost exclusively, so the amount of lighting units really adds up (240 sq/ft).

After searching for various mixes of specialized r/b, paired with white for full spectrum coverage, I happened across the Cree page with their various COB arrays. Upon viewing the spectral output, a higher kelvin range matched the needs of my vegetative needs almost perfectly. For plants that flower, going with the lower kelvin range also seemed to be an excellent match.

For the person struggling with driver output, there are effectively two ways of using constant current drivers. First, if running in series, the rated ma output will be shared across all draws (in this case, the diodes are the draws), and the voltage will be divided. For example, if your driver is rated at 500ma, and 10-30vdc, with 10 diodes, each diode would have 500ma and 3vdc. Alternately, if run in parallel, voltage remains constant, while amps are divided. Take the same circuit as previously mentioned, and every diode would have up to 30v, and have 50ma.

The thing lacking in this thread, that I think is absolutely paramount, is PAR output. While the full spectrum factor outweighs the waste of energy in using white light, at least in my humble opinion, the plants ultimately do have the PAR requirement. As Cree only gives lm output, how does one effectively calculate the size/quantity of diodes required?

Has anyone on here done a PAR test with an assembled unit? Clearly, they work. They are also easy to assemble, and when compared to "retail," they offer a fantastic price point. The flexibility of chip size is outstanding as well for all sorts of applications.
 
The PAR of each of the Cree models has been measured and there are conversion factors (LER and QER) for converting from lumens.

And Fanleaf has an Apogee meter and can probably give you some real-world numbers. :cheesygrinsmiley: Here's a thread:

Anyone have an Apogee AGP-MQ-500 quantum meter?
 
Thanks Graytail, I actually picked up a Apogee SQ-520 last week, and am waiting to build a couple assemblies before doing tests. The new USB model is relatively cheap, and I suffer dearly from a tool obsession. I'll be doing individual tests of some of the smaller COB's, so that I can effectively map things out.
 
So I did find one thread on different forum. One of the members did compare top bin CXB3590, CLU058-1825, CLU048-1818 and VERO 29 v7. Even CLU058-3618 rated for 500W+! In 4000K the Citizen is clear winner and even in 3000K. The 1825 at 90W is as efficient as 3590 at 50W. The 3618 at 300W is running with same efficiency as CXB3590 at 100W. That's something. I know I will be building with either 1825s or 1818s. I need 450w for 2 plants in 4x4 tent but I will keep the girls in the middle. A tough one...
 
So I did find one thread on different forum. One of the members did compare top bin CXB3590, CLU058-1825, CLU048-1818 and VERO 29 v7. Even CLU058-3618 rated for 500W+! In 4000K the Citizen is clear winner and even in 3000K. The 1825 at 90W is as efficient as 3590 at 50W. The 3618 at 300W is running with same efficiency as CXB3590 at 100W. That's something. I know I will be building with either 1825s or 1818s. I need 450w for 2 plants in 4x4 tent but I will keep the girls in the middle. A tough one...
Can i have a link please? I just ordered 2 Vero 29 V7 (3500K,80CRI,79Wx160Lm@2100mA/37x V)
 
Back
Top Bottom