Type of LED

xenia2

New member
Do you guys know if there any articles/threads discuss about the most popular use of DIY LED (brand name)? What are pro and con between different brands?

Thanks
 
No, not quite over. Luexon recently came out with the Luxeon M, which beats every Cree in terms of efficiency and spectrum.

Steve's Leds sells Luxeon M in royal blue and 5k white. $11, but it has the power of four chips in one. Runs at 11.4v at max current of 1000ma. Site is okay, but drivers are somewhat expensive. Meanwells are cheaper. Also, he offers custom 3up stars to minimize disco effect.

Led Group Buy is also a very good DIY led website, but their drivers are very expensive. You can get drivers elsewhere. However, they offer 430nm violets (king for growth and color) and a host of other colors as well (including the Ocean Coral White, which is a quick full-spectrum add-in).
The Maker's Led heatsink is also pretty nice looking.

Rapid Led is the simplest (almost plug and play) but their violets are of a slightly poorer spectrum and efficiency and they don't have an OCW. Drivers are cheap though. And they offer good heatsinks.
If you go the Rapid route, buy neutral white leds and royal blue leds in a 1:2 ratio instead of 1:1 cool white: royal blue. More spectrum to compensate so that you don't need reds and greens.

Choice is yours. Personally, I like Steve's Leds because of the Luxeon M and 3up stars. Then I would get drivers and heatsinks from Rapid, as well as a fan (Steve's sells tube heatsinks, which I don't like).
 
I like the luxeon personally. I ordered them along with steves drivers from steves leds and everything seems to be working fine for me and my set up.
 
So right now the Luxeon's have an edge over Cree's in terms of power utilization, lumens per watt, and price right?
 
Luexon recently came out with the Luxeon M, which beats every Cree in terms of efficiency and spectrum.

That's simply not a factual statement. Comparing the data sheets from Phillips, and Cree, the Luxeon emits in the neighborhood of 800 lm @ .7A and ~11V. Cree's single die XM-L, depending on the bin and color, ranges from 164 - 300lm @ .7A and ~3V. Dependant on the choice of LEDs, you could use 3 of the Crees, vs 1 of the 4 die Luxeons at a 25% reduction in power used.

For a 20 gallon nano that may not make a noticeable difference. On a larger build it probably will.



http://www.datasheets.pl/LEDs/LUXEON-M.pdf

http://www.cree.com/~/media/Files/C...d Modules/XLamp/Data and Binning/XLampXML.pdf
 
Comparing the Cree XM-L and Luxeon M

Comparing the Cree XM-L and Luxeon M

That's simply not a factual statement. Comparing the data sheets from Phillips, and Cree, the Luxeon emits in the neighborhood of 800 lm @ .7A and ~11V. Cree's single die XM-L, depending on the bin and color, ranges from 164 - 300lm @ .7A and ~3V. Dependant on the choice of LEDs, you could use 3 of the Crees, vs 1 of the 4 die Luxeons at a 25% reduction in power used.

For a 20 gallon nano that may not make a noticeable difference. On a larger build it probably will.



http://www.datasheets.pl/LEDs/LUXEON-M.pdf

http://www.cree.com/~/media/Files/C...d Modules/XLamp/Data and Binning/XLampXML.pdf
The Luxeon M neutral white (higher bin, 70 CRI) has a test efficiency(700ma) of 125lm/w, with a typical luminosity of 980lm/w. The Cree XM-L test efficiency is 127lm/w (700ma), with a typical luminosity of 260 lm/w for the 3700k-5000k higher bin.

However, the XM-L tests are done with a temperature of 25*C, versus the test temperature of Luxeon M at 85*C. If we compensate for that (XM-L datasheet has a graph) then it really only achieves about 229lm for an efficiency of 106 lm/w, or 88% of what it is rated for at 25*C. That's only 89% as efficient as a single Luxeon M die.

But one might argue that the XM-L can be run at higher currents, to push out more lumens. However, at a current of 2000ma, almost three times the current used in the above calculations, it can be seen that it only puts out 2.4x more light for a total luminosity of 643lm, or an efficiency of 100lm/w. And compensating for temperature again, we see that it is really only about 566lm, and 89lm/w. About 71% as efficient as a Luxeon M, or 83% as efficient as its 700ma test current.

Cost specs:
At Led Group Buy, a standard XM-L neutral white costs $8, and at Steve's Leds, a Luxeon M costs $12. Both are mounted to PCBs.
At 700ma test current, it takes 4.27(980lm/229lm) XM-L leds to match one Luxeon M, or $34. At 2000ma, 1.76(980/ 556) are needed, or $14. I guess if you ran them at 3000ma, you could almost match a Luxeon M, but then your efficiency would be down the meephole. Not to mention you would get severe heat problems at such a high current to lower efficiency further.

Following this post will compare Luxeon M to XM-L2 leds. The difference is not so staggering there, but give me a few minutes.

XM-L datasheet: http://www.cree.com/~/media/Files/C...d Modules/XLamp/Data and Binning/XLampXML.pdf

Luxeon M datasheet:
www.philipslumileds.com/uploads/354/DS103-pdf
 
Last edited:
Comparing the Cree XM-L2 and Luxeon M

Comparing the Cree XM-L2 and Luxeon M

A Luxeon M from Steve's Leds, mounted to a PCB, costs about $12. A XM-L2 from Led Group Buy, mounted to a PCB, costs about $9. Please tell me if you find a cheaper source, and I will be happy to recalculate.
Let's a take a look at the numbers.

Unlike the previous comparison, both leds are (mercifully) rated for 85*C. So there will be no adjustment needed.

Looking at the 700ma test current on Luxeon M leds, we find they put out 980lm typically, for the 70CRI 4000k bin. That's 125lm/w efficiency.
The 700ma test current on XM-L2 leds, for the T5 bin (3700-5000k), we get 260lm. That's 130.2 lm/w. That's 4% more efficiency than a Luxeon M.

At 1000ma test currents, for the same bins:
Luxeon M: 1323lm, for 118lm/w.
XM-L2: 357lm, for 123lm/w. Still about 4% higher than the Luxeon M.

The Luxeon M costs about 33% more than a XM-L2. It puts out, at a test current of 700ma, 980lm. The XM-L2 puts out 260lm at the same test current that it's rated for.

980/ 260 = 3.77. You need 3.7 XM-L to get the same output of a single Luxeon M.
$9 x 3.7 = $33. A single Luxeon M is $12.

Maybe it's just me, but that 4% higher efficiency just doesn't seem worth the 2.5x increase in cost. I can work out the savings math if anybody wants, depending on the cost per kW in your area.

XM-L2 datasheet:
www.cree.com/~/media/Files/Cree/.../XLamp/.../XLampXML2.pdf‎

Luxeon M datasheet:
www.philipslumileds.com/uploads/354/DS103-pdf
 
Last edited:
ALet's a take a look at the numbers.



Looking at the 700ma test current on Luxeon M leds, we find they put out 980lm typically, for the 70CRI 4000k bin. That's 125lm/w efficiency.
The 700ma test current on XM-L2 leds, for the T5 bin (3700-5000k), we get 260lm. That's 130.2 lm/w. That's 4% more efficiency than a Luxeon M.

XM-L2 datasheet:
www.cree.com/~/media/Files/Cree/.../XLamp/.../XLampXML2.pdf‎

Luxeon M datasheet:
www.philipslumileds.com/uploads/354/DS103-pdf

The numbers skew to your position when you compare the Cree minimum output, the Phillips typical output. Your comparison is close to apples and oranges. Compare the data they both provide, "Min. Luminous Flux @ 700 mA", and the difference becomes almost 3X your calculations, or almost 12% vs. 4% more efficient.

The bottom line IMO is that in the real world you can use 3 - XM-Ls to 1 of the Luxeons for nearly the same effect based on the .7A number. At that current for a build the savings would be approximately 22% in electricity used by the LEDs. If you just have to match the Luxeon output you could bump the Crees to around .8A and the efficiency would still be a little better than 11%.

Today I saw the Crees for $6.50 for 1, Luxeons were $11.00 for 1. That makes 3 Crees cost ~1.8X what the Luxeon costs. Each individual build will vary, but there will be savings, or additional costs for drivers and powers supplies based on the choice of LEDs, all of that will need to be accounted for in the initial cost. If you plan to use the lights you build for three or four years or more the choice is clear.

To put this in perspective, would you rather have a job that pays $12.20/hr or one that pays just $10.00/hr? $11.10/hr or the same $10.00/hr? That's 22% and 11% respectively. On the low end let's say your lights cost $40.00/month to operate, and the LED's efficiency flows through to power consumption, on the low end that's a savings of $4.40/month - over $50 annually for continuing operational costs.

:deadhorse1: Really, in the end, the respective manufacturer's data sheets tell the tale. The Phillips LEDs are probably very good, they're just not as efficient as the Crees.
 
The numbers skew to your position when you compare the Cree minimum output, the Phillips typical output. Your comparison is close to apples and oranges. Compare the data they both provide, "Min. Luminous Flux @ 700 mA", and the difference becomes almost 3X your calculations, or almost 12% vs. 4% more efficient.

The bottom line IMO is that in the real world you can use 3 - XM-Ls to 1 of the Luxeons for nearly the same effect based on the .7A number. At that current for a build the savings would be approximately 22% in electricity used by the LEDs. If you just have to match the Luxeon output you could bump the Crees to around .8A and the efficiency would still be a little better than 11%.

Today I saw the Crees for $6.50 for 1, Luxeons were $11.00 for 1. That makes 3 Crees cost ~1.8X what the Luxeon costs. Each individual build will vary, but there will be savings, or additional costs for drivers and powers supplies based on the choice of LEDs, all of that will need to be accounted for in the initial cost. If you plan to use the lights you build for three or four years or more the choice is clear.

:deadhorse1: Really, in the end, the respective manufacturer's data sheets tell the tale. The Phillips LEDs are probably very good, they're just not as efficient as the Crees.

Cree doesn't list a typical output. Sorry, I didn't notice the word "minimum", as assumed that they were all typical.

Okay, if we use the minimum luminosity, then we get the following:
-Luxeon M gets min. 900lm at 11.2v 700ma, for 114.7 lm/w.
-Cree XM-L2 gets 260lm at 2.85v 700ma, for 130 lm/w.
You get 14% greater effciency. However, if we run the XM-L2 at 1000ma, you get:
-Cree XM-L2 gets 357lm at ~2.92v 1000ma, for 122.3 lm/w.
That's 6.6% higher than a single luxeon M.

You still need, at 1000ma, 2.52 XM-L2 to match one Luxeon M. Where did you find them for $6.50 each? That's integral to the cost.
But anyway, with $6.50 each, you get $5 more to get 6.6% more efficiency.

7.35w of Cree to match 7.84w of Luxeon M. That's about .5w per led. At $0.15 per kw/hr, it will take 5000 hours to save $0.15.
That means it will take 166,666 hours to make up for the discrepancy in cost.


You seem pretty certain that Cree is better, yet you cite the datasheets as the reason. All of these numbers are from the datasheets.

Wait, that can't be right. Hold on...

EDIT: I know there's an error, but I can't find it. Anybody?
 
Last edited:
Oh, loid this discussion again...........

to summerize, BOTH Cree and Luxeon make top of the line state of the art chips as far as what is available to us Reef hobbiests. Each has Pros and Cons depending on your specific design intent but you really can't go wrong with either brand or varrious model.

It all comes down to the design goals of your specific build.

......and just to poke the bear a little bit there is also the consideration of color aesthetics and growth(or PAR potential) of a given LED, not all of the Royal Blues (both brands) are created equal nor are the whites. Some whites boast over 90 CRI while others a measly ~70 CRI. All depends on what you need/want for your application. There will never be a one best chip solution that fits each build. Also a lot comes down to which specific bin is being offered by a vendor, a spec that is not always reliably available, since LEDs are bined by not only wavelenth and color, but also luminous flux/radiometric output, as well as voltage........

:beer:
 
Oh, loid this discussion again...........

to summerize, BOTH Cree and Luxeon make top of the line state of the art chips as far as what is available to us Reef hobbiests. Each has Pros and Cons depending on your specific design intent but you really can't go wrong with either brand or varrious model.

It all comes down to the design goals of your specific build.

......and just to poke the bear a little bit there is also the consideration of color aesthetics and growth(or PAR potential) of a given LED, not all of the Royal Blues (both brands) are created equal nor are the whites. Some whites boast over 90 CRI while others a measly ~70 CRI. All depends on what you need/want for your application. There will never be a one best chip solution that fits each build. Also a lot comes down to which specific bin is being offered by a vendor, a spec that is not always reliably available, since LEDs are bined by not only wavelenth and color, but also luminous flux/radiometric output, as well as voltage........

:beer:

The datasheets list higher CRI chips as well, but many people don't care about CRI IME. So I just used the highest lumen output chips for both in the same color range.

If your design goals are maximum efficiency, then go XM-L2. But the M is, at writing, the most cost efficient.

Why does Vf matter? What are the pros of having a lower voltage and higher current vs. higher voltage and lower current?
 
Cree doesn't list a typical output. Sorry, I didn't notice the word "minimum", as assumed that they were all typical.

Okay, if we use the minimum luminosity, then we get the following:
-Luxeon M gets min. 900lm at 11.2v 700ma, for 114.7 lm/w.
-Cree XM-L2 gets 260lm at 2.85v 700ma, for 130 lm/w.
You get 14% greater effciency. However, if we run the XM-L2 at 1000ma, you get:
-Cree XM-L2 gets 357lm at ~2.92v 1000ma, for 122.3 lm/w.
That's 6.6% higher than a single luxeon M.

You still need, at 1000ma, 2.52 XM-L2 to match one Luxeon M. Where did you find them for $6.50 each? That's integral to the cost.
But anyway, with $6.50 each, you get $5 more to get 6.6% more efficiency.

7.35w of Cree to match 7.84w of Luxeon M. That's about .5w per led. At $0.15 per kw/hr, it will take 5000 hours to save $0.15.
That means it will take 166,666 hours to make up for the discrepancy in cost.


You seem pretty certain that Cree is better, yet you cite the datasheets as the reason. All of these numbers are from the datasheets.

Wait, that can't be right. Hold on...

EDIT: I know there's an error, but I can't find it. Anybody?

No problem, just pointing out a relevant difference with the output comparisons.

Rapid LED has XM-Ls for $6.50.
 
Both Cree and Luxeon are good, solid brands, besides the technology is advancing so fast that what you buy today will be obsoleted tomorrow. As a native Dutchman I support Philips, but I ended up with Cree's :)... (RapidLed didn't have Philips at the time and their office is less than 30 minutes from my office).
 
No problem, just pointing out a relevant difference with the output comparisons.

Rapid LED has XM-Ls for $6.50.

The regular XM-L are less efficient than the M.

But the tech is so fast that Cree will likely develop the "MP-Z" series of leds that are 15% more efficient than Luxeon Mand can be run at 11w at that efficiency.
 
^ Good price on that website.

FizzOut - How do you know which is cool/neutral white, the website doesn't specify.
 
Last edited:
6500K listed in the title indicates cool white
generally:
over 5000k would be cool white
under 3000k is warm white
anything in between would be neutral.
 
Back
Top