DiY LED - Cree CXA3070

CobLedz

Member
COBLEDZ, very cool design! I love the idea of moving side lighting that also provides the circulation (I spend 55W of circulation from a ceiling fan). You didnt ask for constructive criticism but I have to warn a fellow grower, a $6 COB running at full blast is probably less than 15% efficient and the COB will suffer significant lumen depreciation over time and may actually burn out, meaning you would be much better off to use HPS which is 36% efficient. A few tweaks to the design and you could be up to 40% with 3000K LED that will last indefinitely.
Yo Super! Thanks for the kudos. Note to all on riu, please make comments and criticisms. Super? What tweeks do you recommend?
 

SupraSPL

Well-Known Member
Basically just swapping in a CXA3070 Z4 and running at 1.4A with a driver like this. That would bring you up to 39.3% and would put out more light using half the power. You could double up and strap a pair of them to the fan. If you did 4 of them it would match or beat a 400HPS and almost match a 600 HPS.

Also I would double check the heatsink temp. Even though the fan is right there, the heatsink might be in a "dead spot" of airflow? Maybe they would get better airflow if they were strapped to the outer edge.Fan blast.png
 
Last edited:

DiyKindaGuy

Active Member
Glad yu pointed that out supra cause when i first looked at it i thought exactly same thing but didnt speak up thought maybe he had a reason to place it in(deadspot) thumbs up on the redesign...

★★★kushed_out★★★
 

SupraSPL

Well-Known Member
Thatsa good question, although when it really comes down to it only the top bin of the CXA3590 (42.3%) can outperform the CXA3070 Z4 (39.3%). Even if the 2700K or 3000K CXA3590 becomes available, I don't expect the top bin. So the CXA3070 will probably be our best bet for a long time.
 
Last edited:

Bueno Time

Well-Known Member
Ive got my next run vegging under a single 3070 3000k @ 1.4A. Not an ideal light but seems to be vegging better/keeping tighter nodes than with HPS. Had them under 2 42w daylight cfls for first 5 days then put the cob in there. Today is day 15 from seed sprouting.

They are all mixed up in the pics but the strains are 2x CSG Cheese n Chaze and the rest are crosses I made, 3x Blackberry Kush x Chemmy Jones, 3x Pineapple Express x Cheese n Chaze, and 3x Chemdawg 91 s1 x SWT#4 IBL.



Transplant to 1/2 gallon grow bags in 2 days. Got ~3 weeks left until harvest of the plants under 150w HPS, then these babies go in the tent under 2 3070 3000k cobs.
 

DiyKindaGuy

Active Member
Thatsa good question, although when it really comes down to it only the top bin of the CXA3590 (42.3%) can outperform the CXA3070 Z4 (39.3%). Even if the 2700K or 3000K CXA3590 becomes available, I don't expect the top bin. So the CXA3070 will probably be our best bet for a long time.
Maybe cree is just working on a better cob..... most likely end up with two to four cxa3070 inless something else comes out soon

★★★kushed_out★★★
 

PetFlora

Well-Known Member
Equivalent to 4-600 hps?

Mind Blown


Basically just swapping in a CXA3070 Z4 and running at 1.4A with a driver like this. That would bring you up to 39.3% and would put out more light using half the power. You could double up and strap a pair of them to the fan. If you did 4 of them it would match or beat a 400HPS and almost match a 600 HPS.

Also I would double check the heatsink temp. Even though the fan is right there, the heatsink might be in a "dead spot" of airflow? Maybe they would get better airflow if they were strapped to the outer edge.View attachment 3134707
 

mauricem00

Well-Known Member
Basically just swapping in a CXA3070 Z4 and running at 1.4A with a driver like this. That would bring you up to 39.3% and would put out more light using half the power. You could double up and strap a pair of them to the fan. If you did 4 of them it would match or beat a 400HPS and almost match a 600 HPS.

Also I would double check the heatsink temp. Even though the fan is right there, the heatsink might be in a "dead spot" of airflow? Maybe they would get better airflow if they were strapped to the outer edge.View attachment 3134707
are you claiming that 30,000 lumens from an led would provide better growth than 55,000 lumens from an hps lamp? I know that standard hps lamps put out a poor spectrum for growing plants and other light sources come much closer to matching what plants actually use but is there that big of a difference? do not get me wrong I am not promoting HPS. I have never used them. but your statement would suggest almost twice the yield (grams per what) from leds. I grow with T5s and just started a grow with leds but have not seen any comparison grows that would support this claim.assuming 90% driver efficiency thats only 233 watts for four diodes. I am not trying to start an argument. I have not seen data that would support this claim and led grow lights manufacturers have "poisoned the well" with exaggerated claims. also do you know if cree makes a high efficiency cold white led ( 20000-25000K)
 

SupraSPL

Well-Known Member
Fair questions Maurice, it is a complex explanation but the short answer is yes. In this case it would be 208W DiY LED = 27000 lumens versus 400W HPS = 50000 lumens and the LED will still outperform the HPS. (LED=89 PAR Watts, HPS = 140 PAR Watts) Here are the reasons:

-large losses in the HPS reflector, many photons converted to heat
-HPS reflector scatters a portion of the light in the wrong direction
-losses in the HPS fixture's glass, especially as it accumulates dust/smudges
-HPS suffer very significant lumen depreciation
-LED improved spectrum (more blue/red -less yellow/green)
-LED higher efficiency 35% vs 43%
-LED can get much closer to the canopy, more even spread which improves photosynthetic efficiency.

So that all sounds good on paper, and in practice I have been able to replace a bare 600 HPS with ~250W DIY LED of 40% efficiency. When the HPS bulbs were brand new the best I ever got was .5g/W. As the bulbs aged I got .3-.4 G/W consistently. When I started phasing out the HPS the g/W rose to .57 and .67. Now the HPS are completely phased out and I am hoping to get close to 1g/W. Of course these are not high yielding varieties and they are grown in organic soil so I am sure cash croppers can get much higher yields.

I agree LED manufacturers are full of it for the most part. The cheap ones are only 15-20% efficient whereas 600 HPS is 35% efficient, so they would be lucky to match up watt for watt. Of course HPS is so much cheaper so that makes those LEDs a false economy.

Regarding a 25000k CREE, I have never heard of one. Might as well just use a 450nm deep blue since they are up to 55% efficient.
 
Last edited:

tenthirty

Well-Known Member
#1) You can't look at lumen output for a direct comparison. The delivered lumens from an LED will be greater than the delivered lumens from an HPS lamp of similar brightness. The optics on an LED fixture (at least good ones) are precisely aimed for each diode. There is an acrylic or polycarbonate lens over each diode that aims the light on a more direct line. Therefore the efficiency of the optics is higher for LEDs.

#2) Due to the more precise aiming of light, the distribution of an LED luminaire is going to be much more uniform than an HPS area luminaire. The HPS will have a very bright spot (that eats up most of the lumen output) directly below the lamp, which then fades rather rapidly to the edge of the light distribution. The LED luminaire will not have a bright spot, but will be much more evenly lit, spreading the light out. Often, an LED luminaire with half the light output of an HPS will have similar minimum illumination. Average illumination may be lower, but it's more or less irrelevant, since most of that average bump is from the hotspot below.

#3) As mentioned above, the color temperature of the lamp will also produce a whiter light, which makes it easier to discern things. There's a whole range of discussion that can be had on scotopic vs photopic illuminance, but that's a sidebar you don't need to have. Just know that the non-orange light makes it easier to see and grow.

HPS is around 30% efficient.
CXA3070 3000K Z4 @ 1.4A = 52W @ 39.3% efficient.

So IMHO.........yes!!
 

mauricem00

Well-Known Member
Fair questions Maurice, it is a complex explanation but the short answer is yes. In this case it would be 208W LED =27000 lumens versus 400W HPS =50000 lumens and the LED will still outperform the HPS.Here are the reasons:
-large losses in the HPS reflector, many photons converted to heat.
-HPS reflector scatters a portion of the light in the wrong direction
-losses in the HPS fixtures glass, especially as it accumulates dust/smudges
-HPS suffer very significant lumen depreciation
-LED improved spectrum (more blue/red less yellow/green
-LED higher efficiency 35% vs 43%
-LED can get much closer to the canopy, more even spread which improves photosynthetic efficiency.

So that all sounds good on paper, and in practice I have been able to replace a bare 600 HPS with ~250W DIY LED of 40% efficiency. When the HPS bulbs were brand new the best I ever got was .5g/W. As the bulbs aged I got .3-.4 G/W consistently. When I started phasing out the HPS the g/W rose to .57 and .67. Now the HPS are completely phased out and I am hoping to get close to 1g/W.

I agree LED manufacturers are full of it for the most part. The cheap ones are only 15-20% efficient whereas 600 HPS is 35% efficient, so they would be lucky to match up watt for watt. Of course HPS is so much cheaper so that makes those LEDs a false economy.

Regarding a 25000k CREE, I have never heard of one. Might as well just use a 450nm deep blue since they are up to 55% efficient.
thank you. I was not aware that the blues were that efficient. when I switched from standard T5 grow lights to P&A bulbs that peaked at 450 and 658 nm I noticed an increase in yield and considering the spectrum of leds it seems reasonable that they would produce more usable light as well. a 20 to 30 % increase in yield is impressive.I just started some clones under leds and will be interested in seeing how they compare. I am using a cold white led to supplement the blue because the blue leds I looked at did not seem that effective. I am running 1 cold white and 2 warm whites for vegging and will replace the cold white with a neutral white for flowering.if it works well than I will spend the money to upgrade to cree. I am running 650ma on 50 watt cobs ( about 40% maximum rated current) and they still seem very bright. I think DIY is the only way to go at this stage. in time to commercial lights may get better and cheaper but I get the feeling they are just following the DIY crowd. with the price they charge for those things you would think that they could afford to put the best leds available in them.
 

mauricem00

Well-Known Member
#1) You can't look at lumen output for a direct comparison. The delivered lumens from an LED will be greater than the delivered lumens from an HPS lamp of similar brightness. The optics on an LED fixture (at least good ones) are precisely aimed for each diode. There is an acrylic or polycarbonate lens over each diode that aims the light on a more direct line. Therefore the efficiency of the optics is higher for LEDs.

#2) Due to the more precise aiming of light, the distribution of an LED luminaire is going to be much more uniform than an HPS area luminaire. The HPS will have a very bright spot (that eats up most of the lumen output) directly below the lamp, which then fades rather rapidly to the edge of the light distribution. The LED luminaire will not have a bright spot, but will be much more evenly lit, spreading the light out. Often, an LED luminaire with half the light output of an HPS will have similar minimum illumination. Average illumination may be lower, but it's more or less irrelevant, since most of that average bump is from the hotspot below.

#3) As mentioned above, the color temperature of the lamp will also produce a whiter light, which makes it easier to discern things. There's a whole range of discussion that can be had on scotopic vs photopic illuminance, but that's a sidebar you don't need to have. Just know that the non-orange light makes it easier to see and grow.

HPS is around 30% efficient.
CXA3070 3000K Z4 @ 1.4A = 52W @ 39.3% efficient.

So IMHO.........yes!!
I do not use lenses on my experimental led grow light. I have a small grow space and the walls are lined with mylar. the wider beam angle without lenses allows a substantial amount of lite to reflect off the wall and under the canopy with my light 18 inches above the canopy of a test plant the shadows under the canopy disappeared and I was getting good illumination 7 branches down. this method is only useful in small grow cabinets.. regarding the color of light my T5s are purple and the plants love them.
 

DiyKindaGuy

Active Member
I experimented with blue led only in veg... it seemed to me that they were growing slower then the ones with white mixed with blue... also the blue leds made the leaves darker green and the thickness of the leaves was increased as well... so if would suggest only vegging mother plants with completely blue... i was able to keep a mom for six months under two feet with lots of lst though...

★★★kushed_out★★★
 

SupraSPL

Well-Known Member
I should clarify I was not suggesting to use blue only. For veg, I mix the blues in until you get 25-30% (PAR Watts) total blue. My most recent veg lights are warm white, red, deep red, and deep blue. Mixing neutral whites would be great too.
 

SupraSPL

Well-Known Member
Ouch yea that is exactly type of thing I could do. I have blown up a handful of power supplies and drivers due to hooking the 120V to the dc side when I am in a hurry and multitasking.

Did you say you got Z4s in your shipment from Arrow? Those in the pic look like Z2s is why I ask. Z4 on the left Z2 on the right. The Z4 seems to have a thinner layer and you can see the traces much better. Also the Z4 has a smooth surround and the Z2 has a wavy surround.



Anyway, your light bar is lookin badass!
 
Top