DIY LED 220w Cree XTE + Philips ES

FranJan

Well-Known Member
^^Yeah I've read about how ridiculous many manufacturers consider 10,000K+ LEDs, though Cree does make a 10,000K. Manufacturers and their sales departments are just gaming the color standards that were designed long before LEDs hit the scene. It's just another way to make blues to me, and I can't say if they're bad for the plant with out giving it a go. I'm using a 2 x 10 watt 5500K + 1 x 20 watt 12,000K SMD wash lights as cloning/germination lights plus I keep my bonsai mothers under them. They're great, but only up to a point. So maybe B-lux needs to get some better phosphors going, LOL.
 

bassclef

Active Member
Amazing intro Supra, you cover pretty much everything a newbie needs to know. Now I'm going to have to upgrade my Osrams to Crees!!
 

Hosebomber

Active Member
Nothing against you or your research goud, but I'm not trusting a datasheet that has 1 wavelength characteristic for all whites from 2500-25000 and is 3 years old. The chart appears to be for a neutral white at that.
 

guod

Well-Known Member
Nothing against you or your research goud, but I'm not trusting a datasheet that has 1 wavelength characteristic for all whites from 2500-25000 and is 3 years old. The chart appears to be for a neutral white at that.
Osram datasheets sometimes are not really younger.... there is no yearly update!
the chart shows a classical coolwhite spectrum.

the real wavelength characteristic(kelvinpoint)will be given by this chart(here from Cree)
ANSI-XLampXMBL-7.jpg
 

Hosebomber

Active Member
Correct, and if you look at the XY scale, there is no possible way to reach above 10k Kelvin because you drop off the color corrected scale. That is the point where there is no other blending of phosphors and it is simple a blue LED. That is exactly how the tech guys from at Bridgelux explained it to me a few years ago. I believe manufactures that are claiming to have 10k+ kelvin leds are skewing the black body test to appeal to a curtain group, generally coral or aquarium people. I know a number of coral forums that advocate "Get the highest Kelvin rating you can find." These are people that drop from $50-$50,000 on a special fish or rock, so an extra $500-$2000 on the light that claims something greater than all the rest isn't that big a deal.
 

guod

Well-Known Member
Reeducating the improperly educated is harder than correctly educating them in the first place
We Are All Believers
...........................Tao of Physics

Science does not need mysticism and mysticism does not need science, but man needs both
 

SupraSPL

Well-Known Member
Someone else was just asking about the XML. This is the post where I explained why I chose XTE instead.

The XTE and XML whites are both impressive. I came to the conclusion that XTE is better for growing than XML. XTE does have a wider angle but that will improve spectral mixing which allows us to get the LED modules closer to the canopy. The XML cost $6-$10 each while the XTE are $3.25 each. Finally the XTE bins we have available are equally efficient as the XML. Here are the calculated minimum output numbers based on a Tj of 50c and assuming typical Vf. The last column is radiometric efficiency.

View attachment 2438221

XML.png
 

tenthirty

Well-Known Member
Yup, the XTE's look slightly better, but!
Me being a lazy stoner, I would rather solder half the leds
and at 2 to 1 for the XML's the price is a wash.

Just say'in.....
 

SupraSPL

Well-Known Member
If you run the XML at 1500mA it will be 17% less efficient and 20% more expensive than 2 XTE's at 700mA. But I can understand the laziness part to each his own ;)
 

SupraSPL

Well-Known Member
Finally got everything together. Odds and ends are a big part of the job. I settled on prolimatech pk2 thermal paste to mount the stars to the heatsink. Prolimatech is more expensive than Ceramique but it does have better thermal transfer. $32 for 30 grams.

For connecting strings, heatsinks and drivers a cheap and fool proof solution has been 1/4" crimp on slide connectors with heat shrink tubing. Female for positive and male for negative so it is impossible to mix them up and drivers are easily interchangeable. It also makes it easy to bug in a multimeter to measure current. It is surprising how many connections you end up with. I went through 100 connectors and ordered 200 more $12. 15 feet of heatshrink tubing $8.

$10 for a 1LB roll of 60/40 lead solder. $10 for 150 feet of copper tape. $15 for 500ft of 1mil thin kapton. $15 for 50 ft of jack chain. Good prices to stock up on and makes the job a lot smoother.
 
Yeh Xt-e are nice too, i believe XMl has a wider spectrum all around. You can get the Xm-l for $6.50 a piece, ledgroupbuy.com I run the xml's around 1800-2000ma, which is recommended by cree, but it can go up to 3000ma depending on cooling, but not much or a difference in lumen output. So i would say 2000ma is best, which is around 742lm a chip. You would need two Xt-e running at 1500ma to match that highest bin 856lm(full blast).

also Xm-l's run at a average Forward voltages of 3.4, So the only down side is you would need twice as many drivers or (driver voltage) and watts for the 2 XT-e's(3.4v x2 =6.8v) to match 1 Xm-l. Dependent on what you prefer. They are both really good chips. Just glad we can share info here, DIY is the way of the future...for sure
 

green surfer

Well-Known Member
Nice thread Supra,

i'm glad to find it and see that knna's spirit is still alive!

I agree that efficient warm whites are the way to go. Keep up the good work!


EDIT: Supra,

I don't think that the best bin actually available for XTE warmwhite is R3, i found only R2 in the datasheets.

The bests XTE bins for warmest tints actually available at Cutter are Q5 8A (2700K) and R2 6A (3500K-3750K).

Even a R2 warmwhite is a very good bin, and Cree are binned at warm temp (85°C)

A Luxeon rebel or a Dragon is binned at 25°C.

Stanley has efficient warmwhites too, they don't do the chips, but they have very good phosphor layers, this is a Japanese company.

the GSPW1653JTE-30X seems to be interesting, 140 to 150lm, and there is no need to use 450nm leds IMAO.

Capture d’écran 2012-12-21 à 12.52.08.pngCapture d’écran 2012-12-21 à 12.47.19.jpgCapture d’écran 2012-12-21 à 12.48.05.jpgCapture d’écran 2012-12-21 à 12.51.30.jpgCapture d’écran 2012-12-21 à 13.01.01 - copie.jpgCapture d’écran 2012-12-21 à 13.03.51 - copie.jpg

http://www.stanley-components.com/en/search/search_top.cfm?select_product_class=STC001
 

SupraSPL

Well-Known Member
Cutter has the XT-E 3000k in R3 in the 10mm round size. They list it as R2 but they have lots of typos on their site so I go by the part number: Led Part Number is XTEAWT-0-7D2-R30-FL-0001. This is a 3000K Colour Bin. Just above that they list this one: Led Part Number is XTEAWT-0-5D2-R30-FL-0001. This is a 5000K Color bin. There is a typo above that says it is an R4 bin. Also, 5D2 is not 5000K it is 4000K. Moral of the story is that I have to depend on the part numbers they give and hope they don't have typos in there too lol.


I am not sure the XTE binning and labeling sheet is complete, Luxeon sheets have the same problem. For example the XTE sheet shows the R3 bin as the maximum available at 4250K and then R5 as the maximum available in the 4300K. Bypassing the R4 bin completely which makes no sense. Either way, good catch green surfer! Hopefully what I received is actually the R3 7D2.
 
Top