DiY LED - Cree CXA3070

captainmorgan

captainmorgan

Well-Known Member
Someone around here might be able to calculate the numbers for you and tell you the percentages,that's beyond my skill level.
I'm not sure what the magic number is,I've just seen the symptom with a strain that I know well and the only thing that was different was the amount of 660 I added.
 
SupraSPL

SupraSPL

Well-Known Member
I agree with Cap regarding the ratio. I always keep 630 in excess of 660. Although the 3000K CXAs do peak around 610, there is still a lot of output in the 630nm range and most of the power in the lamp is coming from the COBs, so you can go heavy on the 660 and still end up with most of your red in the 630 range. Either way that is a badass setup!

Regarding the 3000K Vero 10, it outperforms the 3000K XML2 which I had never considered before. At .25A the Vero 10 dissipates 6.3W at 35% efficiency and cost $0.76/W. At 2A the XML2 dissipates 6.4W at 30% efficiency and cost $0.73/W. So for shorter canopies it makes plenty of sense to use the Vero 10. For even lower light requirements (vegging) you could run Vero 10 softer .115A and it still slightly beats the XML2.
 
Positivity

Positivity

Well-Known Member
What I've noticed running through some red combos..

240w white + 80w deep red = nicest buds but too much red to leave on in stretch phase

240w white + 40w red + 40w deep red = not quite as nice buds

So I won't bother playing with red 630. And I won't go as high as 80w deep red in a 2 x 4 area.

I think the magic is in neutral/warm white with just a small amount deep red. Still working on the amount I like...taking it down to 30w in a 2 x 4 this time.

730 at lights out to trigger sleep time, and 660 during on hours to trigger dense flowers

Just a theory and subject to change at any second
 
Bueno Time

Bueno Time

Well-Known Member
SupraSPL said:
I agree with Cap regarding the ratio. I always keep 630 in excess of 660. Although the 3000K CXAs do peak around 610, there is still a lot of output in the 630nm range and most of the power in the lamp is coming from the COBs, so you can go heavy on the 660 and still end up with most of your red in the 630 range. Either way that is a badass setup!

Regarding the 3000K Vero 10, it outperforms the 3000K XML2 which I had never considered before. At .25A the Vero 10 dissipates 6.3W at 35% efficiency and cost $0.76/W. At 2A the XML2 dissipates 6.4W at 30% efficiency and cost $0.73/W. So for shorter canopies it makes plenty of sense to use the Vero 10. For even lower light requirements (vegging) you could run Vero 10 softer .115A and it still slightly beats the XML2.
Click to expand...
Positivity said:
What I've noticed running through some red combos..

240w white + 80w deep red = nicest buds but too much red to leave on in stretch phase

240w white + 40w red + 40w deep red = not quite as nice buds

So I won't bother playing with red 630. And I won't go as high as 80w deep red in a 2 x 4 area.

I think the magic is in neutral/warm white with just a small amount deep red. Still working on the amount I like...taking it down to 30w in a 2 x 4 this time.

730 at lights out to trigger sleep time, and 660 during on hours to trigger dense flowers

Just a theory and subject to change at any second
Click to expand...
Thanks for the input guys.

I should probably (hopefully) be ok then since my WW COBs are putting out the vast majority of the light and the 10 660nm is only ~15w total between the two bars. Also since I will leave the bars off for the first 3 weeks or so that should help avoid the issue of causing more stretch, which I am constantly combating in this small space.

Ill give it a shot and see most likely.

Also good info on the Veros vs XML2s, does that also stand for the 4500kish range or just the 3000k.
 
guod

guod

Well-Known Member
it´s not the red that makes the trouble here, you simply run out of blue in the mix.

Try this...
go to http://www.bmlcustom.com/custom-led-strip/

place 6 warmwhite(3000K)
take a look at the spectrum and check the amount of blue -->10%

now add one red
check the spectrum again and the amount of blue --> 8%
repeat until blue is --> 5%
this is also the minimum blue, plants need. HPS has the same problem .
blue wls.jpg
 
Last edited:
Bueno Time

Bueno Time

Well-Known Member
guod said:
it´s not the red that makes the trouble here, you simply run out of blue in the mix.

Try this...
go to http://www.bmlcustom.com/custom-led-strip/

place 6 warmwhite(3000K)
take a look at the spectrum and check the amount of blue -->10%

now add one red
check the spectrum again and the amount of blue --> 8%
repeat until blue is --> 5%
this is also the minimum blue, plants need. HPS has the same problem .
View attachment 3160074
Click to expand...
Thanks for chiming in Guod.

Yeah, I have messed with that spectrum analyzer thing before. The thing is I cant get the ratios of 3000k to 660 630 450 right, there arent enough spots on the "bar" that you put your led choices on. If I put 7 3000k on there, I can put 1 660 on and be about right ratio wise for power but then I only would have half a dot worth of 450 blue and 630 red. Since I cant do half a dot worth of a wavelength to make the ratio proper I would then need to put 14 3000k circles on the bar, 2 660, 1 circle of 450 and 1 of 630 to be close ratio wise and there isnt enough spots for that.

Not sure if that makes any sense to you guys but it makes sense in my head.

SupraSPL said:
That is true, might want to put a few blues in with your deep reds. I added Luxeon ES blue to my newest modules for the same reason.
View attachment 3160085
Click to expand...
I based my R/B bars off of your recommendations for supplementing R/B when running large amounts of WW and the pic you just attached, I actually spent almost a half hour looking for that post with the pic yesterday before I finally found it. I am running my COBs harder than those on your fixture (more WW) and also added 1 more 450 on my R/B string. So theoretically if that ratio of WW to red and blue is working out fine for you without issues of bud stretch, then I should be good too I would think, since I would have more WW from running the COBs harder and 1 more blue in my fixture.

Edit: I forget how many of those red are 660 and how many are 630 on your fixture in the picture. Was it 3 630s 3 450s and the rest are 10 660?

If it is then ya my WW is running harder and I added same amount of 660, 1 more 450 and 1 more 630 on my R/B string.

SupraSPL said:
BT, good question about the neutral white Vero 10s vs XML2 for vegging, I will look into that.
Click to expand...
Thanks, I will be waiting for a spreadsheet or conclusion. :)
 
Last edited:
Bueno Time

Bueno Time

Well-Known Member
SupraSPL said:
Sadly this 2.1A driver delivered nowhere near what it promised. The useable vF range is 20-30 so it is no use for our COBs. Efficiency was 85-87% and it is power factor was corrected.
Click to expand...
I love when you are the guinea pig and test stuff out for us. You spend your money on questionable items just to test them for the greater good of the LED community. Hats off sir.
 
SupraSPL

SupraSPL

Well-Known Member
Haha tnx BT, I enjoy it. So I compared the Vero 10 5000K (22% blue) and the XML2 U2 4500K (28% blue) and they are about equal in terms of efficiency and price when running soft (2-3W). If run harder (6W) the Vero 10 wins.

The downside of the Vero (for me) if that in order to run soft we need a driver putting .115A. The lowest output driver I have is .27A.
 
happy75

happy75

Well-Known Member
Supra, have you already taken a look at the newest bin of the cxa3070, the AB? Is that just the old Z4 or is it truly better than the Z4 regarding par and/or efficiency? 80 pcs of the AB are offered to me at an very affordable price, going to decide on monday if I will buy them. I only need 40 or so..
 
MrFlux

MrFlux

Well-Known Member
guod said:
???
a slightly warmer color temp(4500K / 5000K) contains 6% more blue.(28%/22%)

no, not really
Click to expand...
Cree really tends to have a lot more blue. Even the 3700K Cree has more than the 5000K Bridgelux:

XML2 3700K
420-480 nm blue : 24.4% power, 20.0% photons

XML2 5000K
420-480 nm blue : 30.2% power, 25.3% photons

Vero 3000K
420-480 nm blue : 8.1% power, 6.3% photons

Vero 4000K
420-480 nm blue : 12.6% power, 10.0% photons

Vero 5000K
420-480 nm blue : 22.3% power, 18.1% photons
 
C

CannaBare

Well-Known Member
Can I ask what your room atmosphere temperature is? I shoved 16 CXA's in a 4x4 and its getting up to 86 without my UVB lights on.

I had two designs and I came across something that stumped me. If I have one heatsink with air flowing half speed and another flowing at full speed. If the heatsinks have an equal constant temperature over a 12hr period. Will the heat dissipated into the air technically be the same? It a hard question for me to ask too I guess.

Thanks!
 
SupraSPL

SupraSPL

Well-Known Member
happy75 said:
Supra, have you already taken a look at the newest bin of the cxa3070, the AB? Is that just the old Z4 or is it truly better than the Z4 regarding par and/or efficiency? 80 pcs of the AB are offered to me at an very affordable price, going to decide on monday if I will buy them. I only need 40 or so..
Click to expand...
I have a test in the works to compare Z2, Z4 and AB. That is awesome that you have access to the ABs! Any chance you can get 2700K Z4s?
 
SupraSPL

SupraSPL

Well-Known Member
Cannabare how hard are you running them? I am dissipating 500W in a 7'X8' tent with a canopy of 4'X5" and canopy temps are just a few degrees above ambient. I intend to expand that to as much as 800W by increasing the ventilation fan speed if necessary.

Yes I believe that both heatsinks would dissipate the same amount of heat over time, but the temperature differential across the the heatsink will be larger with the fan at half speed. So the junction temp of the COB would be much higher.
 
You must log in or register to reply here.
Top