DiY LEDs - How to Power Them

[Observe & Report]

OK! I was confused about test current. All Vero models have the same efficiency at test current. Maybe all the CXAs do too, I didn't check. However, there is no correlation between the XML-2 and CXA3070 test currents.

So I took the relative flux vs current charts from both datasheets and scaled them to be the same then overlaid them. The X axis (current) is aligned but I translated the graphs slightly in the Y axis (relative flux) so we could directly compare the curves of XML-2 vs CXA3070 at 85C, which is the only Tj plotted on the XML-2 chart.

*XML-2 is black and CXA3070 is green




I was surprised to see these curves match up so closely when the current is aligned. So I recalculated the efficiencies of both parts at 800mA.


Cree CXA3070 80-CRI 3500/3000K bin Y4 @ 800ma 85C = 98.2 lm/W

Cree XML-2 80-CRI 2600-4300K bin T2 @ 800ma 85C = 98.7 lm/W


At 800ma, the CXA3070 and CREE XML-2 have the same efficiency. I would guess that is because they use the same diode and phosphor tech. When you look at the chart, however, you can see that as the current goes up the CXA3070 starts losing. Therefore, the assertion "Multis aren't as efficient at creating light as a COB " is false. Apparently, the benefits provided by COB tech do not include efficiency.

For our purposes, COBs are superior to multis for all of the other reasons that SupraSPL mentioned (cost, assembly, intensity.) The exception would be where you can't have much distance between the emitters and plants, such as PC grows or side lighting.


This is how I learn and I just got schooled, thank you!

 

[stardustsailor]

660nm is where it's at. Look at the CX2540s 93cri 3000k..couldn't imagine the 2700s with 93 CRI. http://www.cree.com/LED-Components-and-Modules/Products/XLamp/Arrays-NonDirectional/XLamp-CXA2540. Wish I could tell you efficiencies.

You should check also the rel.power of 700-730 nm ...

And that note comes from someone that uses High CRI ( 95% ) 2700 K WW leds ...(OSRAM OSLON )..
Having very similar -relatively to other WW ,almost 'same'- REL. Power spectral graph,like those Crees...
Plenty of them ....

Great led to use ,from middle flowering till end ....

Before ,you really do not want to have plenty of them switched on ...
No-no-no-no...
Few,maybe ...
I'm not still sure ...

 

[Scotch089]

660nm is where it's at. Look at the CX2540s 93cri 3000k..couldn't imagine the 2700s with 93 CRI. http://www.cree.com/LED-Components-and-Modules/Products/XLamp/Arrays-NonDirectional/XLamp-CXA2540. Wish I could tell you efficiencies.

Wow.

THAT makes me want to DIY.. best looking white I've seen. Where do we get them? Do we know the eff#'s?

 

[stardustsailor]

OK! I was confused about test current. All Vero models have the same efficiency at test current. Maybe all the CXAs do too, I didn't check. However, there is no correlation between the XML-2 and CXA3070 test currents.

So I took the relative flux vs current charts from both datasheets and scaled them to be the same then overlaid them. The X axis (current) is aligned but I translated the graphs slightly in the Y axis (relative flux) so we could directly compare the curves of XML-2 vs CXA3070 at 85C, which is the only Tj plotted on the XML-2 chart.

*XML-2 is black and CXA3070 is green

View attachment 3027490


I was surprised to see these curves match up so closely when the current is aligned. So I recalculated the efficiencies of both parts at 800mA.


Cree CXA3070 80-CRI 3500/3000K bin Y4 @ 800ma 85C = 98.2 lm/W

Cree XML-2 80-CRI 2600-4300K bin T2 @ 800ma 85C = 98.7 lm/W


At 800ma, the CXA3070 and CREE XML-2 have the same efficiency. I would guess that is because they use the same diode and phosphor tech. When you look at the chart, however, you can see that as the current goes up the CXA3070 starts losing. Therefore, the assertion "Multis aren't as efficient at creating light as a COB " is false. Apparently, the benefits provided by COB tech do not include efficiency.

For our purposes, COBs are superior to multis for all of the other reasons that SupraSPL mentioned (cost, assembly, intensity.) The exception would be where you can't have much distance between the emitters and plants, such as PC grows or side lighting.


This is how I learn and I just got schooled, thank you!

Therefore, the assertion "Multis aren't as efficient at creating light as a COB " is false. Apparently, the benefits provided by COB tech do not include efficiency.

At last ....
Someone....

Spread ...
Both of light and heat ...
What a tiny diode emits must spread as much as possible ...
(in a grow situation,at least ....)
Heat and light ..
That is the real advantage !!!
(For home scale horticultural use ...)

Concentrate the heat and light of many into one "bad-@$$ -single source of light ?
When one has the real chance -at last- getting to spread the light AND the heat losses !!!

Why one would that in his/her grow-room ?
Really why ?
For what reason ?

Yes I do care also for my wallet/pocket ...
And the simplicity ..
And the time ..

But hey ...

 

[SupraSPL]

That is very interesting that they have the same efficiency at a given current, thanks for pointing that out. I did not realize that the CXA suffers from slightly higher current droop at high currents. Luckily that does not come into play because we run them much softer than maximum. COBs also have a much better thermal resistance than SMDs, so if you did run them hard they would suffer from less temp droop which would more than make up for the slight difference in current droop. This same thing is noticeable when comparing the red XPE and XPE2.

However, if we compare the 3000K top bins of each at 700mA, you can see that the CXA3070 is much more efficient again. 150lm/W for the Z4 bin CXA3070 vs 137lm/W for the T5 XML2. To add insult to injury, the XML2 cost $3.12/W and the CXA cost $1.51/W. To take it up a notch, the CXA3590 top bin in 3000K makes 158.5lm/W for $2.38/W (@350mA). So COBs really are kicking some booty and that is just what we growers need.

Regarding the distance, I keep the COBs about 8" above the canopy, they are ideal for running bare.

 

[Observe & Report]

Why one would that in his/her grow-room ?

"penetration"

Very intense, concentrated light gives you greater distance from the source when the top of the plant is at optimal flux. That means the lower parts of the plant are relatively closer to the light then they would be if the tops were closer. That means the lower parts are getting more light.

For example, you set things up things so the tops are getting optimal light 12" away. 12" down the plant, now 24" away from the light source, you're getting a lot less light because you're twice as far away. However, if your lighting is much more intense and the tops are best placed 24" away then 12" down from the plant you're 36" away from the light source, which is only 50% further. So the buds 12" down are getting a lot more light than the ones on the other plant, even though the tops are receiving the same amount of light. Keep increasing the light intensity and the optimum distance and the "window" of good lighting increases with it. When you get to 93,000,000 miles away, the bottom of the plant will receive the same amount of light as the top of the plant, even if it is 12 feet tall.

 

[stardustsailor]

That is very interesting that they have the same efficiency at a given current, thanks for pointing that out. I did not realize that the CXA suffers from slightly higher current droop at high currents. Luckily that does not come into play because we run them much softer than maximum. COBs also have a much better thermal resistance than SMDs, so if you did run them hard they would suffer from less temp droop which would more than make up for the slight difference in current droop. This same thing is noticeable when comparing the red XPE and XPE2.

However, if we compare the 3000K top bins of each at 700mA, you can see that the CXA3070 is much more efficient again. 150lm/W for the Z4 bin CXA3070 vs 137lm/W for the T5 XML2. To add insult to injury, the XML2 cost $3.12/W and the CXA cost $1.51/W. To take it up a notch, the CXA3590 top bin in 3000K makes 158.5lm/W for $2.38/W (@350mA). So COBs really are kicking some booty and that is just what we growers need.

Regarding the distance, I keep the COBs about 8" above the canopy, they are ideal for running bare.

A good and fair point ...
But I am allowed to like the multiples ,am I not ?


Another thing I dislike with COBs is that you can not "match' them (regarding light cones ) so good (like the multiple minis (,
even in a small array ...

Still ,you made a good point...

 

[stardustsailor]

"penetration"

Very intense, concentrated light gives you greater distance from the source when the top of the plant is at optimal flux. That means the lower parts of the plant are relatively closer to the light then they would be if the tops were closer. That means the lower parts are getting more light.

For example, you set things up things so the tops are getting optimal light 12" away. 12" down the plant, now 24" away from the light source, you're getting a lot less light because you're twice as far away. However, if your lighting is much more intense and the tops are best placed 24" away then 12" down from the plant you're 36" away from the light source, which is only 50% further. So the buds 12" down are getting a lot more light than the ones on the other plant, even though the tops are receiving the same amount of light. Keep increasing the light intensity and the optimum distance and the "window" of good lighting increases with it. When you get to 93,000,000 miles away, the bottom of the plant will receive the same amount of light as the top of the plant, even if it is 12 feet tall.

LOL!
..............

 

[PetFlora]

Multiple small cobs ~ 10-20w, maybe 30w (or 50w?)seems like the ideal as they cover ~ 12"- 24" square footprint, possibly throwing a very nice 2-3 ft umol footprint

Here the cobs could be a mix of CW/NW/WW with separate o/o switches

As to penetration, there are several Defoliation techniques that would significantly open up the lower 10-20" of a plant

 

[SupraSPL]

A good and fair point ...
But I am allowed to like the multiples ,am I not ?

You can have both. Just finished this module this morning. There are 3 XPE reds P3 bin, 5 Luxeon ES deep reds EX6 bin, 5 Oslon SSL80 deep reds 3T bin, 3 Luxeon ES royal blue M4R bin all on one string, will run it at 650mA. The COBs will be running at 900mA (32W each) so total dissipation for this module will be 89 Watts and efficiency will be 41.4%. Total blue output will be 15% (PAR Watts). There is a 140mm fan on the back of the heatsink it will be running at 8.4V.

 

[stardustsailor]

I like it ...
The heatsink seems superb ...
The led placing ..Could have been a bit ...different ....
Blue disperses a lot,still it's blue diodes we are talkin...
Better to have them as far as possible..Maximize their distance apart ...
Like in a triangle shape ....

Warm white light is heavy on middle towards long wls ...'Directive' light and not dispesing omindirectional like short wls do ...
As far apart as possible ...

Let the reds fill rest of space left ....


Superb ...
( Although I'd used couple of NW 5000 K COBs instead of the blue leds )...

Still I think It is of 'High-end" grow efficiency ...

Same 'color' school as .. ..
LOL!!!
Ha-ha-ha...
Oh,what's with your blues ???

( Kidding...Ya know,I dislike blue leds ... )

Both great led DIYers / designers ...
You and him ...
"RBW school .."

 

[SupraSPL]

Ya it had 4 COBs on the heatsink originally and I had to choose 2 to remove. Then I just worked with the space I had left so it would have better spacing if I had started from scratch. No worries though itll get the job done.

Those blues are 56% efficient at 650mA. That is why I use them instead of cool whites.

 

[SupraSPL]

Wow.

THAT makes me want to DIY.. best looking white I've seen. Where do we get them? Do we know the eff#'s?

Scotch, you have probably seen this chart, adjusted by Mr Flux to show true relative curves. Suffice it to say, efficiency is not good as far as diy goes and the deep red advantage is very small. In this case, you would be much better off with the "more photons" option.
View attachment 3027794
In the Cree datasheet the just put the highest peak of each curve at 100% which really makes the high CRI look deceivingly awesome.

 

[FranJan]

http://www.ebay.com/itm/100W-Cree-XM-L2-LED-Multichip-100-Watts-White-Light-10000LM-Copper-Plate-/360731477320?pt=LH_DefaultDomain_0&hash=item53fd458548

I just can't imagine cooling something like this and getting anywhere near the performance you can out of each XML2. Supra what do you think of this design? I had 3x10 XPE whites and I had to use active cooling to run them at 900mA, though I should of had bigger heatsinks.

 

[RedCarpetMatches]

Scotch, you have probably seen this chart, adjusted by Mr Flux to show true relative curves. Suffice it to say, efficiency is not good as far as diy goes and the deep red advantage is very small. In this case, you would be much better off with the "more photons" option.
View attachment 3027794

Don't you use deep and far red?

 

[SupraSPL]

I agree FJ, that is a less than optimal design, although it is very convenient. It demonstrates the concept of why COBs shouldnt be awesome and why I was very skeptical of COBs at first.

 

[stardustsailor]

Scotch, you have probably seen this chart, adjusted by Mr Flux to show true relative curves. Suffice it to say, efficiency is not good as far as diy goes and the deep red advantage is very small. In this case, you would be much better off with the "more photons" option.
View attachment 3027794

Dear Supra allow me ,to offer my experiences and approach with these leds ...
Well yes,efficiency wise they do suffer ..
In veg they will definately stretch the plants and will favor stem tissue formation (also 'favoring' carbohydrate production :lignin/cellulose/etc )
and will negatively impact Nitrogen assimilation while-at the same time- they push plants to consume a lot of stored Nitrogen
( in vacouoles as nitrate ions ),
"burning up " /"drying" - scorching the leaves ...Specially along with 660 nm red leds ...


I do not like them ...

But ...

I've to say/admit that ....

Utilising them after every possible "internodal stretching" period (after the 3rd or so week of flowering ) ...
Oh ....Superb flowering ..Thickness-Swelling-Maturing..All it's there with that extra FR that they have ...

Still I do not like them...

But in flowering they do wonders ...

But because they are unefficient ,they have to be used in certain pieces and above ..
Then,if so,FR becomes pretty considerable power-wise ...
Tricky ...

Lots of cons ..
But superb pros ....

 

[SupraSPL]

Yes they will grow some badass nugs I don't doubt that I just want to make sure potential DIYers dont get misled by that Cree chart

 

[stardustsailor]

Yes they will grow some badass nugs I don't doubt that I just want to make sure potential DIYers dont get misled by that Cree chart

Yes ...
You have the Good ones ....
The Bad ones ...
The Ugly ones ...

And the Deadly ones ( Led Fatale ) ...

So alluring ...So sexy ..So destructive if ...abused ....

 

[Observe & Report]

However, if we compare the 3000K top bins of each at 700mA, you can see that the CXA3070 is much more efficient again. 150lm/W for the Z4 bin CXA3070 vs 137lm/W for the T5 XML2. To add insult to injury, the XML2 cost $3.12/W and the CXA cost $1.51/W. To take it up a notch, the CXA3590 top bin in 3000K makes 158.5lm/W for $2.38/W (@350mA). So COBs really are kicking some booty and that is just what we growers need.

Ouch. I see my mistakes. Cree has a minimum efficiency for a certain generation of tech before the part is scrapped, so if you pick the bottom bin from two different parts they'll have the same efficiency. I understand the thermal and mass production advantages of COB but what is it that makes it more efficient when Tj is constant?

Don't think I'm a COB hater! I just recall you or Flux talking about how the XML-2 is a bit more efficient but not worth the expense or effort. I couldn't find the post then naively did the calculations and ran with it.