Space requirements for 600w of CREE CXA 3070

Hey everyone,

I've decided to bite the bullet and go for an LED grow. I'll be using the much beloved CXA 3070 COB (10) with the typical meanwell drivers @ 1.4a - so i'm looking at just under 600W of light (~589w?). I know it's not the most efficient setup (lumens/watt wise) but the high price of each COB+driver+heatsink means i need to work them a little harder to make it worth my while.

I was originally planning on putting everything in a 4'x 4' tent, but i've been advised this might be overkill. I know COBs get hot (but not as hot as HID), but the AC should manage that quite well (9k BTU), I'd have thought?

Should i upgrade to a 5' x 5' tent for the 10 COBs or stick to my 4'x 4'? I'm starting to think a 5' x 5' may be better (looking to grow 14 ladies, aero clone/veg, flower in NFT), as i expect to be mainlining with 8 main colas and this will increase my footprint.

Any input is much appreciated!

My plan is 16@1050a(38w each) for a 1000w replacement(4x4+)

So no I don't thinks it's over kill for the plants. Heat I dont know but let's hope there is a good difference and it's should be runnable.
I used to run 600hps in a 3x3 tent and was fine. I know that 600w of led will run pretty cool comparatively to a 1000hps.

I would go with Nine 3070's for a 4x4. Doing a 1' 4 " grid or spacing would Cover a 4x4 very nicely.

Greengenes707 said:

My plan is 16@1050a(38w each) for a 1000w replacement(4x4+)

So no I don't thinks it's over kill for the plants. Heat I dont know but let's hope there is a good difference and it's should be runnable.
I used to run 600hps in a 3x3 tent and was fine. I know that 600w of led will run pretty cool comparatively to a 1000hps.

Click to expand...

I agree with this, you should be fine OP even in a 3x3 with your 9k ac.............Damn GG your building a 600w 16 cob panel?!?

nft for flower.........hmmm, hope you have a level/patience with it=== not much room for error IMO.

Thanks for the input guys,

I was going for NFT for the flower as in StinkBuds guide (a NFT/ebb&flood hybrid) ... but i'll probably make change the design so it's more of a 3x3 grid than a row. Admittedlty, it will be my first grow, so it's probably going to go wrong. Things will happen i can't fix this time, but it's all part of the learning process. Luckily, everyone here seems really nice and able to help!

I've heard keeping on top of NFT systems is tricky, but i'll be using RO water, i'll have a spare pump and be checking the nutes/ph twice a day (morning/night). I'll have a screen in place so they don't fall, etc.

I was hoping to mainline, but a 4x4 tent with 9 plants there might not be enough space. I'm looking to pull 2 zips per plant which is ~1GPW, which should be do-able with CO2.

I know I'm biting off more than i can chew, but hey, that's half the fun!

Thanks again

Since this is your first, I'd advise running something simple like lucus formula. In RO water the pH stays within spec(5.5-7.5). I rarely ever check my pH. Very stable formula with stable growth.

Look at gaius diy perfect example of a diy in my opinion

Sent from my SAMSUNG-SM-G900A using Rollitup mobile app

According to Mr. Flux's numbers in http://rollitup.org/t/cree-cxa-analysis.743645/ the 80CRI 3000K CXA puts out 1.57 uMol/s per Watt at nominal current. That's about 927 uMol/s at 589W. You have about 1.5 square meters of space. That means the average PPF in your tent will be around 624 uMol/s/m2. Most people recommend more like 800 uMol/s/m2 for flowering. You'll get more or less output dependening on the chosen color, bin, and the drive current efficiency relative to whatever it is for the CXA1304 at the test current but it sounds like you need more PPF.

Thanks for that Observe & Report!

I'm a bit of a numbers geek, so that's gold to me. I'll be running the 3070's, not the 1304 ... i'll do some digging and see what the PPF is for my setup, but it looks like it'll only be a couple of COBs away in the worst case!

Thanks, you're a life saver!

For 10 cxa3070 3000k z4 in a 4x4 I get;

- 707.24 PPFD @ 1.05A
- 817.54 PPFD @ 1.40A

When deciding what multiple to run, it's best to start at the driver level IMO. If you will run individual drivers for each COB then this is a moot point.

The highest efficiency choice that's been tested in this forum for 1.05A operation is the meanwell HLG series. They work out to be pretty cost effective as well and offer a great warranty (piece of mind for me). You can run 5 COBS on this driver.

If I had a target wattage of about 600W for a 4x4, I would run 3 strings on 5 @ 1.05 (1060 PPFD @ 570 Watts, 603 Watts at the outlet after driver loss).

I opted for 707 PPFD at 380W (402 Watts at the plug) in a 2 strings of 5 @ 1.05 for a 4x4 since I plan to run 2 plants only (personal medicine, not a cash cropper). In my opinion this is around the best part of the photosynthetic efficiency curve to be in given design and hydro costs. This is to say, if I were to run more wattage, I wouldn't increase intensity but instead add more light over a larger area. Some other members here do share the same philosophy.

If you plan to run that many plants I would imagine they won't get that tall, and perhaps you could consider 15 of the cxa in a 5x5 at 1.05A? That would be a hard setup to beat for the cost and running expenses.

Observe & Report said:

According to Mr. Flux's numbers in http://rollitup.org/t/cree-cxa-analysis.743645/ the 80CRI 3000K CXA puts out 1.57 uMol/s per Watt at nominal current. That's about 927 uMol/s at 589W. You have about 1.5 square meters of space. That means the average PPF in your tent will be around 624 uMol/s/m2. Most people recommend more like 800 uMol/s/m2 for flowering. You'll get more or less output dependening on the chosen color, bin, and the drive current efficiency relative to whatever it is for the CXA1304 at the test current but it sounds like you need more PPF.

Click to expand...

Sounds like an awesome build! Since you are going with a relatively large setup, you might benefit from using large heatsinks rather than CPU coolers. One of the cheapest profiles is the 4.85" from heatsinkUSA and fits a 120mm fan. The 5.886" fits a 140mm fan. That would allow you to use less and larger fans and would simplify the wiring/installation. The downside would be less adjust-ability.

I have run into a problem with the Chinese "50W" drivers from eBay. The first batch I got tested awesome, 1.4A and 89% efficient. But I just got a batch that puts out only 700mA at 37vF. It puts out 1500mA at 25vf and is only 83% efficient no matter which vF. I am working on a refund and started sourcing drivers elsewhere.

The Meanwell LPC-60-1400 is a good option but only 85% efficiency which is surprising and not power factor corrected. The HLG-185H-C1400A is a good high voltage option like the one OTIMT posted above. It is good for only 3 COBs but priced not much higher than the LPC-60-1400 and much higher efficiency. It is power factor corrected. The downside is introducing high voltage into the grow space.

http://www.powergatellc.com/mean-well-hlg-185h-c-power-supply.html

10 CXA3070 Z4s at 1.4A would dissipate 520W at 39.3% efficiency (205 PAR W). If I am understanding Mr Flux data correctly, any 3000K CXA produces 4.88umol/S per PAR Watt. So 204 PAR W = 996umol/S. In a 4X4 that is 12.75 PAR W. sq ft or 607 PPFD. I use 10 PAR W/sq ft in my canopies and have no complaints on bud size. If anything I need to increase my fertilizer to keep them from fading early.

SupraSPL said:

The HLG-185H-C1400A is a good high voltage option like the one OTIMT posted above. It is good for only 3 COBs but priced not much higher than the LPC-60-1400 and much higher efficiency. It is power factor corrected. The downside is introducing high voltage into the grow space.

http://www.powergatellc.com/mean-well-hlg-185h-c-power-supply.html

Click to expand...

Hey Supra, how do you think 3 of the 3070 Z4 3000k's would do for a 30"x30" cabinet with one of these HLG's?

Another question,the Ideal COB holders can use up to a 18 gage wire,would that work for 3 Z4 COB's at 1.4?

Hmm thatsa good question Cap, that would be 61.3 PAR W for 6 sq ft, 10.2 PAR W/ sq ft. That is right where I am at and it works well.

For comparison, if you went with 5 COBs on the HLG-185H-1050A as OTIMT mentioned you would get 84 PAR W = 14 PAR W/sq ft. (I calculated at 1150mA because that driver actually puts out up to 1150mA).

This is a very interesting question. Mr Flux posted this;

CXA 3000K
Power in : 3.8 W
Luminous flux : 423 lumen
Efficacy : 111 lumen/W
LER : 344 lumen/W
Radiometric eff.: 32.3%
Radiant flux : 1.23 W
Photon flux : 5.98 umol/s
Blue : 10% power, 8% flux
Red : 24% power, 27% flux

My interpretation was that the radiometric eff was simply radiant flux / power consumed. 1.23 / 3.8 = .323

My thinking is that, at this particular efficiency, it produces 5.98 umol/s at 3.8 input watts (32.3%, 1.23/3.8 )

Say I were to run a much larger 3070 COB at 1.05, would not the radiometric efficiency scale from the original cxa analysis based on the 1304?

So basically what I did was determine the difference in efficiency that I would be running at, and apply that correction factor.

It is very likely that I am way over my head in all of this. Just when I thought I had a grasp on it! Time to brush up on my terminology as that may have led to my misinterpretation of Flux's data.

What do you think Supra?

SupraSPL said:

10 CXA3070 Z4s at 1.4A would dissipate 520W at 39.3% efficiency (205 PAR W). If I am understanding Mr Flux data correctly, any 3000K CXA produces 4.88umol/S per PAR Watt. So 204 PAR W = 996umol/S. In a 4X4 that is 12.75 PAR W. sq ft or 607 PPFD. I use 10 PAR W/sq ft in my canopies and have no complaints on bud size. If anything I need to increase my fertilizer to keep them from fading early.

Click to expand...

You are correct to divide dissipation power by radiant flux to get radiometric efficiency. At some point Mr Flux updated his data slightly based on using a different curve. These are the numbers I was working with, which I was under the impression are the newest numbers:

CXA 3000K
Power in : 3.76 W (9.40V x 0.40A)
Luminous flux : 423 lumen
Efficacy : 112 lumen/W
LER : 328 lumen/W
Radiometric eff.: 34.3%
Radiant flux : 1.29 W
Photon flux : 6.29 umol/s
Blue : 10% power, 8% flux
Red : 24% power, 27% flux

Also worth mentioning, Cree claims the LER is 325, a close agreement with Mr Fluxes calculation of 328.

So the way I figured it, 1.29 PAR W = 6.29umol/S so 1 PAR W = 4.88 umol/S. So the next challenge is to figure out the radiometric efficiency of the situation in question, which is kind of tricky because you have to have an idea of Tj/Tc. Once you know the radiometric efficiency you can multiply by dissipation wattage to get PAR W, and then just multiply by 4.88 to get umol/S (ppf) emitted. For that I recommend using either the Cree Product Characterization Tool or this chart:

Yes you would get more umols/ dissipation W, but you would get the same umols/ PAR W.

Sorry Supra, I deleted my post because I replied before your edit.

Hmmm, I definitely have some more reading to do then!

It would appear that my reasoning was incorrect, although I don't completely understand why. Off to wiki!

Sincerely appreciate the replies, as I'm sure many others will.

It may be confusing because the chart says "power in" where I call it dissipation. I usually reserve "input power" to refer to how much the LED and driver draw from the wall. Dissipation power is the LEDs light + heat. That is where radiometric efficiency comes into play, to separate the light from the heat. Then you get PAR Watts, which is just the light in the visible/photosynthetic spectrum, photons expressed in Watts or umol/S.

The 4.88umol/S figure applies to any Cree CXA3000K curve, independent of the radiometric efficiency. It is simply a way to convert PAR W to umol/S that is actually weighted depending on the energy of the photons, which varies by wavelength. That number would be different for every color temp or SPD curve and it is thanks to Mr Fluxes awesome calculus that we have it I hope that helps clear it up

All that said, if you find that MY reasoning is incorrect or my interpretation of his work is incorrect, that would not surprise me LOL.

Right before you posted the reply above I had the "Ahhhhhaaaa!" moment. I was under the belief that umol/s were independent of the radiometric efficiency which is why I was applying a correction factor. Your last post reinforced my understanding of the concepts at play and set me straight, so to say!

So I'm, looking at 787 umol/S in my 4x4, or about 530 PPFD.

Lower than I had estimated but still enough for about a 22.5 DLI on a 12 hour cycle, or almost 46 DLI 24/0.

Supra, I can't thank you enough for the information you supply to all of us DIYers