Gonna do my first DIY LED with Cree CXA3590

REALSTYLES

REALSTYLES

Well-Known Member
Just purchased 4 Cree CXA3590 3000k COB's. I already own some Mars II LED's but now I see LED lights work, I'm stepping up the game. In a few threads I was being a dick to guys who say Chinese lights suck but I say differ. The Cree COB's I just bought are the best LED's right now and made in China. My goal is to save money on growing so I've gone LED. I thought about building one 4 years ago and I'm glad I waited. I'll keep you guys posted as soon as I get my first run of COB's
 
FrozenChozen

FrozenChozen

Well-Known Member
mc130p said:
Nobody ever said all chinese made diodes suck. Good luck with your project.
Click to expand...
I think he misunderstood the statement.... Chinese LED fixtures (which use lower quality diodes) are garbage compared to COB technology... Chinese panels DO NOT STAND UP TO THEIR CLAIMS. That's the bottom line. On top of that, the diodes are driven to the point that quick failure is all but imminent. Most flowers produced under strictly cheap "Chinese" panels are not sub par in quality, just quantity...
 
REALSTYLES

REALSTYLES

Well-Known Member
FrozenChozen said:
I think he misunderstood the statement.... Chinese LED fixtures (which use lower quality diodes) are garbage compared to COB technology... Chinese panels DO NOT STAND UP TO THEIR CLAIMS. That's the bottom line. On top of that, the diodes are driven to the point that quick failure is all but imminent. Most flowers produced under strictly cheap "Chinese" panels are not sub par in quality, just quantity...
Click to expand...
It actually depends on what Chinese light you buy. But this thread is about Cree CXA3590 and if you wanna tlak China LED's are garbage go to this link and we can discuss how they suck https://www.rollitup.org/t/led-noobs-stop-buying-china-junk.828872/page-26#post-11147618
 
REALSTYLES

REALSTYLES

Well-Known Member
Just ordered 4 more CXA3590 but 5000k and Supra I'll be hitting you up for a little help on the right drivers to run the beast lol.
 
REALSTYLES

REALSTYLES

Well-Known Member
I give a special thanks to Digi-Key gonna get my heatsinks from heatsinkusa.com and figure out what fans to use with fan drivers. I still love my Mars II but I can do better than Apche tech and Blackdog and Area51 for lesser than I paid for my Mars II.
 
SupraSPL

SupraSPL

Well-Known Member
Edit: LMAO I was typing this post up at the same time you asked.

I think you will be impressed by the efficiency and lack of heat. What bin did you get (the BB bin from Digikey?)

Do you have drivers picked out? The CXAs excel at being driven soft while the Vero29 excels at being driven hard. But if you do drive the CXA hard, they will still grow some serious nugs. Here are a few drivers that I know will work for CXA3590

300mA, $3 (drives 1 CXA3590)
300mA, $7 (drives a pair at 280mA)
300mA, $8 (plastic shell, should drive a pair @ 280mA)
350mA, $56 (drives 3, dimmable)
500mA, $70 (drives 5, dimmable) HLG-185H-C500A
700mA, $70 (drives 3, maybe 4, dimmable) HLG-185H-C700A
1050mA, $70 (drives 2, dimmable) HLG-C185H-C1050A

The HLG-120H-C series would work, $64 ea.

This charts shows how the CXA and Vero compare at each drive current. If this chart is mumbo jumbo, just focus on the dissipation wattage, efficiency and cost columns and feel free to ask a lot of questions :leaf:
CXA3590 vs Vero29.png

For example, at 1050mA, they both cost the same/photon, but the CXA is dissipating 80W and the Vero only 38W. The CXA is 36.7% efficient and the Vero 42%. There are a few caveats though, from my testing the Vero draws more power than their charts suggest and the CXA binning is more precise, the Vero just lumps all the different bins into one grab bag so you might get significantly more or less than "typical" performance from them.
 
Last edited:
REALSTYLES

REALSTYLES

Well-Known Member
SupraSPL said:
Edit: LMAO I was typing this post up at the same time you asked.

I think you will be impressed by the efficiency and lack of heat. What bin did you get (the BB bin from Digikey?)

Do you have drivers picked out? The CXAs excel at being driven soft while the Vero29 excels at being driven hard. But if you do drive the CXA hard, they will still grow some serious nugs. Here are a few drivers that I know will work for CXA3590

300mA, $3 (drives 1 CXA3590)
300mA, $7 (drives a pair at 280mA)
300mA, $8 (plastic shell, should drive a pair @ 280mA)
350mA, $56 (drives 3, dimmable)
500mA, $70 (drives 5, dimmable) HLG-185H-C500A
700mA, $70 (drives 3, maybe 4, dimmable) HLG-185H-C700A
1050mA, $70 (drives 2, dimmable) HLG-C185H-C1050A

The HLG-120H-C series would work, $64 ea.

This charts shows how the CXA and Vero compare at each drive current. If this chart is mumbo jumbo, just focus on the dissipation wattage, efficiency and cost columns and feel free to ask a lot of questions :leaf:
View attachment 3313396
Click to expand...
First order CXA3590-0000-000R00BB30F-ND, second order CXA3590-0000-000R0HCB50F I hope I got top bin. Damn $64 ea. for the HLG-120h-c would I need eight of them?
 
SupraSPL

SupraSPL

Well-Known Member
For 3000K, the top bin (and the most efficient warm white LED that was ever for sale) is the CB although currently sold out. The next best is BD, also sold out. The BB can be considered a mid bin and of course Digikey tends to get a premium when it comes to CXAs but they are the only game in town most of the time.

For the 5000K, CD is the top bin for now(the most efficient cool white that was ever for sale) and the CB is the next best bin.
CXA3590 vs Vero29 5000K.png
 
Last edited:
SupraSPL

SupraSPL

Well-Known Member
Basically, if you plan on driving them harder than 300mA, you would have to go with the HLG series (unless anyone knows of some generic high voltage drivers above 300mA) But if you go with the HLG-185H-C700A/B, you could probably get away with a pair of drivers, although it may have to be dimmed to 600mA to stay under the Vf limit of the driver (286Vf). So that works out to about $20 per COB, not too bad for a 94% efficient driver :) Your total dissipation would be about 370W and efficiency about 42% for the 3000K, 49% for the 5000K.
 
Last edited:
REALSTYLES

REALSTYLES

Well-Known Member
Damn Digi-Key charged me California sales tax and their in Mn. I called them them and they said they had to because of Nexus which they voluntarily collect the tax and send it back to Ca. That's the first time I've been charged tax from out of state. Sometimes living in Cali sucks all they want to do is bleed you out of money. I should be receiving my Cree's by Wednesday.
 
nogod_

nogod_

Well-Known Member
I have hlg-120h-c500a in hand with the intent of running 4x cxa3590s @ 500ma. Will post thread once finished.

I also ordered the hlg-185h-c700a with the intent of running 4 x cxa3590s @ 600ma and hoping the string doesnt exceed the drivers voltage limits. Its on back order thru jameco until the 21st.

If you want to wait until january to see the results you can do that, im just waiting to get back to the shop/drill press.

The 120h (150w max) driver will definitely work but i want to see if i can squeeze every last drop of juice from these suckers.

Good luck!
SupraSPL said:
Basically, if you plan on driving them harder than 300mA, you would have to go with the HLG series (unless anyone knows of some generic high voltage drivers above 300mA) But if you go with the HLG-185H-C700A/B, you could probably get away with a pair of drivers, although it may have to be dimmed to 600mA to stay under the Vf limit of the driver (286Vf). So that works out to about $20 per COB, not too bad for a 94% efficient driver :) Your total dissipation would be about 370W and efficiency about 42% for the 3000K, 49% for the 5000K.
Click to expand...
 
Greengenes707

Greengenes707

Well-Known Member
I have 10 hlg185-1400's and 2 120-700's. All run great...and even with some shorts and wiring fuck ups, they are working perfectly and un-damaged. I also have the eBay drivers that work great also...but no dimming and just one cob per and run warm for how efficient they are. Supra just did a good post elsewhere showing that ~5% increase in drive efficacy can be the same as gaining a whole bin or more.
 
You must log in or register to reply here.
Top