DIY Modular LED-Light

next spectrum tool >>> BETA TEST

http://www.ledspecs.com/


Flower-Spectrum of the new Light with it


Veg Spectrum...

-Guod ,I've some indications (just indications ,though...) ,that some yellow -amber light ( 580-620 nm ) is needed for flowering ...

From the HPS era ,I've always used Philips 400 W GreenPower ( Best of the best IMO ...)...
For both vegging and flowering ..
With great results, up to 7-8 full grows,per lamp ....
A lamp with great amounts of yellow/ambers ....


I'm already testing this thing ....
Maybe I'm wrong ....
But those wls ,have something to do with flowering ....

( I've been reading/studying a bit about yellow-amber wls , different species of insects and how they react to those wls and
the relation of all these with flowering plants ....Something is there ....I'm not sure yet .... )

stardustsailor said:

-Guod ,I've some indications (just indications ,though...) ,that some yellow -amber light ( 580-620 nm ) is needed for flowering ...

Click to expand...

same here...
the new light fills this Range with some Cree XPG WW/CW( peaks at 610/550nm)
the amount in this Range is about 40% in Veg. Mode and 20% in Flower mode.
the Cree Leds are working on a fixed current, so the PPfD(in µmol)is the same in in both/all settings

I think you're building another great light ,there ...
Keep up the great work .

very clean cut project. quality work buddy. is this a kit you purchased or is this from scratch?

mehrific said:

very clean cut project. quality work buddy. is this a kit you purchased or is this from scratch?

Click to expand...

from scratch.

under the merkaba-light, for now 10 days
beam angle 60° - distance 12cm

10 days at flowering you mean ....
Great looking plant ,Guod !

Guod ,a question .....

While Thin Film Tech leds for sure are way more efficient than Volume Emission Leds ,as
also they have really low R Th ( thermal resistance ) ......

But is it possible for Volume Emission dies-chips , to be more heat tolerant ,due to thicker materials used in die/chip construction* ?

*( Thicker semiconductor "sandwich " )

Know anything about it ?

I'm asking ,cause those cheap ,dam' ,asian leds ( thick sandwiched dies ) ,seem really to withstand lots of mis-use ,
especially high working temps ....

For sure are ....of third choice ....
But for the ones lacking money and/or the experience needed with high quality leds
and what comes along with them ( high quality ...cooling & driving ,I mean .... ) .....
Then those cheapos ,might seem a good idea ...At least something to start with and gain some experience ...

But really ....
Brother they can work (24x like on Astir panels ) only with the mcpcb ....
With no heatsink at all ..

(Ok ,probably lifetime will suck and actual operating output is rather low due to high Tj ... )

...
But is it possible to withstand more heat "abuse " ,cause of thicker dies ?

stardustsailor said:

Guod ,a question .....

While Thin Film Tech leds for sure are way more efficient than Volume Emission Leds ,as
also they have really low R Th ( thermal resistance ) ......

But is it possible for Volume Emission dies-chips , to be more heat tolerant ,due to thicker materials used in die/chip construction* ?

*( Thicker semiconductor "sandwich " )

Know anything about it ?

Click to expand...

no clue!

i know there are some structure as
planar: very old
TIP(Truncated Inverted Pyramid) >Lumileds?
Aton/Nota >OSRAM
SIC: >CREE
Thin film: >OSRAM
flip chip: >Lumileds

http://www.nusod.org/nusod06/Frontpage_files/ThB1.pdf

but what is inside a chinese Led...???

but what is inside a chinese Led...???


I'll tell you the few I'm aware of ....

Before some time (can't recall when exactly ..A year maybe ....Two years ? )
Anyway ...
Some patents from Bridgelux ,regarding some of their "old " dies (MCXB or something ..can't recall ,really ..A whole one or two series ),expired ....

So B/L sold the rights or something like that ,to plenty of small -Asian triangle ( PRC -Taiwan - HK ) -manufacturers...

So now they are making their own "Brigeluxx " blue dies (As for reds ,they 've been making them long time now .... ) ....
Dies based on 'old' B/L 's designs ....
Exact same die ,even on-die wiring design ...(Checked by macro photography and B/L's die pdf ..Exact same ! )
So ...But they are pretty thick dies ..(volume emitters ) ....
They "sit "in the center , of a parabolic mirror 'cup'/'disc' * ..
*(To reflect side emission upwards )

.... Soft Silicone encapsulation ,above cup & die ..
And then the outer PMMA protective "lens " ....

Die & cup sit directly on a copper heat slug (pretty thick ) ..
So heat slug is not out-of-circuit / "dead" ,due to direct contact of die's base with top surface of heat slug ...
(mirror 'cup' has a hole in the center to fit the die ..)

Easy to make ,bulky & usually shitty construction...
(bubbles in silicone,mis-aligned dies -phosphor gaps -dull mirror 'cups' -scratched pmma outer lenses -etc ) ....

Nothing to do with high quality ceramic leds ....

Oslons fit inside a PMMA lens of a cheapo led ....That difference of dimensions .....

But they withstand a lot of mis-handling and abuse ....

Can't reflow them easy ....

Case and lens are made of thermo-plastic materials and if you follow i.e. Oslons reflow soldering profile ....
At half of procedure ,they've already turned into a white/transparent blob of melted plastic ....

And you know what ?

Dam' things ,still work afterwards ...

Nice pdf ..Lot's of info ...
Lot's of studying tonight ....

Love the LED threads on rollitup! After looking through some threads, most of the guys seem to be using Cree WW LEDs, especially XT-E, my question is, what's the way to go here, more efficient Cree emitters or better spectrum with the Oslons? Do you have experience with this? (esp. guod and SDS)

Some examples:

3000K Oslon (CR7P.EC-KTLP-5R8T-1-350-R18 )




2700K 95CRI Oslon (CQ7P.CC-JUKQ-5U8X-1-K)




XM-L2 Chart(s)



The Osram parts have a very high forward voltage and 800mA maximum current which kind of limit their usability. Cree are good to go with a bigger die size, more efficiency all the way but lacking in the higher red department.

tebos said:

Love the LED threads on rollitup! After looking through some threads, most of the guys seem to be using Cree WW LEDs, especially XT-E, my question is, what's the way to go here, more efficient Cree emitters or better spectrum with the Oslons? Do you have experience with this? (esp. guod and SDS)

Click to expand...

our next Light will have the best of both worlds. cree XPG(2) for CW and WW, Osram for red, blue and WW.
first ideas...
http://s1073.photobucket.com/user/guvoo/library/yabo

yabo

Nice, waiting for your results!

I just looked through the datasheet of the 95-CRI Oslon and compared it with a Nichia 119 2700K 90-CRI, the spectral distribution is almost identical (the Nichia has more blue but the reddish graph is the same). A lot of deep red seems to be the key to a high CRI, it's fair to guess that the Cree XM-L2 and XP-G2 90-CRI will have the same spectral distribution with a lot of red in the 650-700nm range. Looks like you can ditch the Oslons altogether and use the much more efficient Cree counterpart!

tebos said:

Nice, waiting for your results!
...
Looks like you can ditch the Oslons altogether and use the much more efficient Cree counterpart!

Click to expand...

much more efficient Cree...


we use the cree leds for the wider Beam angle, to get a better overall mix of the spectrum.

The Osram WW-Leds with the smaller beam angle are placed on the edge.
as they are on the red Channel, less Blue is here also a goody!

XM-Ls should be even more efficient. But what I wanted to say, is that the yellowish (2,7k-3k) High-CRI LEDs probably have all the same spectral distribution (good color rendition is achieved by the same colors being dominant, doesn't matter which company), peaking at 625nm and ranging all the way down to 700nm. Crees datasheet isn't accurate here because they only show graphs with average values for a whole product line (different bins and tints).

I don't want to spam in your thread any more, will post in the https://www.rollitup.org/led-other-lighting/602112-diy-led-grow-lights-experiences-25.html thread.

guod. really great work...
after i saw the model of your m50 , i tried exact the same but with 1w LEDs, im wondering if its gonna work out ? (i used 2x13 1w leds, mine became 26w totally)

i had a 12v switching power from before (10A max).. so got leds series in four by four groups... and connected groups in parallel...