LED Spectrum Customization Question

[unlmtd216]

Greetings Rollitup forums! I am in need of some knowledgeable advice having to do with customization of spectrum and proper placement of diodes within a grow light panel to correctly balance different colours. I am currently working with a LED company in China in order to introduce their product within North America. I have done some research in the field but I am no expert...at the moment anyway.

I ordered a sample unit and requested a wavelength ratio chart to see what kind of colours to expect. I did research and found that currently, the unit isn't fully capable of putting out the correct spectrum for the benefit of cultivating Cannabis, which is what I specialize in. I made some changes to the current wavelengths and have requested the company for feedback on the new wavelengths I require. I would appreciate the opinion of others who are learned in the area of LED grow lights...

Here are the original wavelengths and my requested changes (in the brackets):

410NM 1PCS (replace with 425nm)

430NM 2PCS (replace with 425nm)

460NM 9PCS (replace with 445nm)

505NM 2PCS (replace with 475nm)

585NM 2PSC (replace with 475nm)

630NM 14PCS (replace 7 with 640 and 7 with 660nm)

660NM 24PCS (replace 12 with 675nm, 6 with 450nm, and 6 with 480nm)

730NM 2PCS (no change)

6500K 2PCS (no change)

12000K 2PCS (no change)


So the new wavelength quantities would be:

425nm x3
445nm x9
450nm x6
475nm x4
480nm x4
640nm x7
660nm x14
675nm x7
730n x2
6500k x2
12000k x2

Here's a layout of the placement of the individual diodes:



Would this be considered as an optimal 8-band spectrum LED? The blue to red ratio is about 1:1 so I figured it would serve for the entire life cycle of the plant. Also, How does the placement of the colours look? Could they be improved in any way? Any advice would be greatly appreciated. Thanks Guys!

 

[littlejacob]

Bonjour
What kind of diodes this pannel use
What is the binning of diodes
What kind of driver's using in it, how are they efficient? how much Amp they deliver and how much real watt?
Heatsink and fans size?
That is what you should be aware of...
Have a great day ★

 

[littlejacob]

Oh, yes the company name or a pic of your pannel will help us a lot, but i am pretty sure that I already have an idea of this pannel shape...
If you already bought it, i hope it was really cheap...

 

[alesh]

Greetings Rollitup forums! I am in need of some knowledgeable advice having to do with customization of spectrum and proper placement of diodes within a grow light panel to correctly balance different colours. I am currently working with a LED company in China in order to introduce their product within North America. I have done some research in the field but I am no expert...at the moment anyway.

I ordered a sample unit and requested a wavelength ratio chart to see what kind of colours to expect. I did research and found that currently, the unit isn't fully capable of putting out the correct spectrum for the benefit of cultivating Cannabis, which is what I specialize in. I made some changes to the current wavelengths and have requested the company for feedback on the new wavelengths I require. I would appreciate the opinion of others who are learned in the area of LED grow lights...

Here are the original wavelengths and my requested changes (in the brackets):

410NM 1PCS (replace with 425nm)

430NM 2PCS (replace with 425nm)

460NM 9PCS (replace with 445nm)

505NM 2PCS (replace with 475nm)

585NM 2PSC (replace with 475nm)

630NM 14PCS (replace 7 with 640 and 7 with 660nm)

660NM 24PCS (replace 12 with 675nm, 6 with 450nm, and 6 with 480nm)

730NM 2PCS (no change)

6500K 2PCS (no change)

12000K 2PCS (no change)


So the new wavelength quantities would be:

425nm x3
445nm x9
450nm x6
475nm x4
480nm x4
640nm x7
660nm x14
675nm x7
730n x2
6500k x2
12000k x2

Here's a layout of the placement of the individual diodes:



Would this be considered as an optimal 8-band spectrum LED? The blue to red ratio is about 1:1 so I figured it would serve for the entire life cycle of the plant. Also, How does the placement of the colours look? Could they be improved in any way? Any advice would be greatly appreciated. Thanks Guys!

If I were to build a LED lamp from small diodes I'd go with 1:1 ratio of 630nm red and white (CCT depending on the output of the red LED, probably 4000K-5000K).
But it's 2015 and the COBs rule now.

Regarding your lamp - more than half of the power is in the blue region, certainly not suitable for a flowering lamp. Might work for vegging but I'd reduce the blue ratio even for that.

Most importantly - be warned. You're most probably going to get some not-very-good LEDs. You can put 25 bands into it if you want, but you won't get anywhere with inefficient LEDs.

 

[unlmtd216]

Bonjour
What kind of diodes this pannel use
What is the binning of diodes
What kind of driver's using in it, how are they efficient? how much Amp they deliver and how much real watt?
Heatsink and fans size?
That is what you should be aware of...
Have a great day ★

Thanks for the reply...here are the specs of the model...there are pics at the bottom of the individual chips/drivers. They are all 100% CREE brand chips so, this is not a cheap unit.

Size 394x218x62.5mm
Actual Power 120w±10%
Input Voltage AC100~240V/50~60Hz
Output Current 630mA
Output Voltage DC60~90V LED Quantity 60pcs
LED Type 3w Cree Chip Shell Material Al6063 +PC
Working ‐20° ~40° ，45%~95%RH Color Gray
Lenses 60/90/120 Degree
Driver 2pcs Isolated Driver
Fan Double roller and low noise Features Salt Fog and UV Protection
Lifespan 50,000 h
3 Years Warranty
Certification CE RoHS N.W. 3.3g

 

[unlmtd216]

If I were to build a LED lamp from small diodes I'd go with 1:1 ratio of 630nm red and white (CCT depending on the output of the red LED, probably 4000K-5000K).
But it's 2015 and the COBs rule now.

Regarding your lamp - more than half of the power is in the blue region, certainly not suitable for a flowering lamp. Might work for vegging but I'd reduce the blue ratio even for that.

Most importantly - be warned. You're most probably going to get some not-very-good LEDs. You can put 25 bands into it if you want, but you won't get anywhere with inefficient LEDs.

Okay...so how about some pointers on wavelengths? What should they be to have a lamp that would serve in both areas of growth; veg and flower? Why would you choose 4k-5k white light instead of the blue spectrum? I read blue 425nm and 675nm are the best. Here are some quotes:

"Most horticultural scientists agree that the highest amount of usable wavelengths of
the spectrum occur at red (675 nm) and blue (425 nm)? Yet, these two frequencies occur
least in MH and HPS bulbs currently on the market. Hmmm… Not to mention newer models,
such as the LED UFO promises as much lumen output as regular 400-watt MH/ HPS lamp –
but using only 90 actual watts. Very impressive." -quote - http://www.hightimes.com/read/let-there-be-light

"(425 nm) This wavelength helps to support peak plant chlorophyll absorption during the
growing process. Blue light promotes photosynthesis and encourages vegetative growth
and dense plant growth during the vegetative grow stage." - quote - http://www.growbigorgrowhome.com/hqrp-135-watt-led-grow-light-wavelengths-that-support-vegetation-and-bloom-grow-big-indoors-with-leds/

 

[littlejacob]

Bonjour
I don't know if the spectrum is ideal...but the pannel look very nice and well made compare to some I saw (if all the pannel look like the small part you showed) and the black color is so cool.
How much did you pay for it and how did you arrive at this spectrum (why did you make this customisation, who inspired you?)
Have a great day ★

 

[alesh]

Okay...so how about some pointers on wavelengths? What should they be to have a lamp that would serve in both areas of growth; veg and flower? Why would you choose 4k-5k white light instead of the blue spectrum? I read blue 425nm and 675nm are the best. Here are some quotes:

"Most horticultural scientists agree that the highest amount of usable wavelengths of
the spectrum occur at red (675 nm) and blue (425 nm)? Yet, these two frequencies occur
least in MH and HPS bulbs currently on the market. Hmmm… Not to mention newer models,
such as the LED UFO promises as much lumen output as regular 400-watt MH/ HPS lamp –
but using only 90 actual watts. Very impressive." -quote - http://www.hightimes.com/read/let-there-be-light

"(425 nm) This wavelength helps to support peak plant chlorophyll absorption during the
growing process. Blue light promotes photosynthesis and encourages vegetative growth
and dense plant growth during the vegetative grow stage." - quote - http://www.growbigorgrowhome.com/hqrp-135-watt-led-grow-light-wavelengths-that-support-vegetation-and-bloom-grow-big-indoors-with-leds/

Cool whites have a lot of blue and beside that, they're the most effective source of light in the green-yellow region. Basically it's red power, and some white to add some blue and broaden the spectrum.

BTW The first quote is ridiculous.

 

[unlmtd216]

Bonjour
I don't know if the spectrum is ideal...but the pannel look very nice and well made compare to some I saw (if all the pannel look like the small part you showed) and the black color is so cool.
How much did you pay for it and how did you arrive at this spectrum (why did you make this customisation, who inspired you?)
Have a great day ★

Yea, it is a nice quality product made by a good company. The spectrum isn't final yet...I am trying to get tips from experts...seems there aren't many on this forum in the area of LEDs...I just did some reading really all over the internet on cannabis websites...like the ones I quotes.. and forums basically... I am aiming for a well balanced light spectrum that will promote peak chlorophyll production.

Cool whites have a lot of blue and beside that, they're the most effective source of light in the green-yellow region. Basically it's red power, and some white to add some blue and broaden the spectrum.

BTW The first quote is ridiculous.

hmm..interesting..thanks for the info.

Any other opinions on what an efficient and balanced light spectrum would be?

 

[Stephenj37826]

What cree leds are they using? Also what bin?

 

[unlmtd216]

What cree leds are they using? Also what bin?

I believe they are Cree XPE 3535

 

[gk skunky]

Trying to make a mono based blended light is like designing a robust diet plan for people. Works better for some than others. While using quality whites is like just maintaining a better life style, balanced diet already included. LOL

 

[Stephenj37826]

I tend to lean twords cool white with added reds. Look at apache tech. Maybe a little uv/ir in there as well. Unless there is a big jump in 3/5 watt diode efficiency cob will be top dawg for a while? Remote phospor is the next big thing as well.

 

[Meinolf]

Last week I got my hands on a sample of a 20W 7blue/7red/6white XPE light on a PCB in a kind of standard floodlight housing. The company's rep thought this light was the shit (as you could also dim the channels individually), but I actually couldn't help but laugh. It still made me question my assumptions, since his company was investing in blurple lights in 2015. Sure, he was correct clorophyll has absorption maxima in the red and blue region. I then double-checked Taiz et al. 2015 to find that chlorophyll still absorbs all other wavelengths within 400-700, that the maxima are identified by irradiating single pigments in vitro and that the factual absorption is substantially affected by the chloroplast structure (e.g. multiple layers), and that the photochemical quantum yield is approx. 0,95. So I am glad it still comes down to umol/J, as long as the spectrum fairly resembles the McCree curve. I wondered why companies still invest in gloomy blurples, when an autodidact layman like me can already debunk that myth, but I see there is still quite a demand on the uneducated consumer side, sometimes even in "professional" horticulture.

 

[unlmtd216]

I tend to lean twords cool white with added reds. Look at apache tech. Maybe a little uv/ir in there as well. Unless there is a big jump in 3/5 watt diode efficiency cob will be top dawg for a while? Remote phospor is the next big thing as well.

I appreciate the info...What wavelengths ratios do you use ? and what kind of cool whites? 12000k? Could you possible give me some percentages of each type within a 3wx60 light? Thanks.

 

[nevergoodenuf]

Go buy 2 Cree cxb 3070's, 2 alpine 11 plus cpu coolers, and a 100 watt driver. The XPE chips can be anywhere from 60 to 120 lumen per watt. Most likely it will be closer to 60 lumen per watt. They aren't going to source high bin chips, unless you specify what you want. At these outputs you might as well use cfl's. The crappiest CXB 3070 bin @ 50 watts will get you over 110 lumen per watt minimum and best bin @ 50 watts is over 170 lumen per watt.

 

[stardustsailor]

I'm not sure if this helps ...
Maybe in China still the all-mighty "magic spectrum " remains unknown ...
but for sure ,in Taiwan ,they know plenty more ,than the chinese .......

https://www.google.gr/url?sa=t&rct=j&q=&esrc=s&source=web&cd=42&cad=rja&uact=8&ved=0CB0QFjABOChqFQoTCMey5sL14scCFQZtFAodeNwH-Q&url=http://www.mdpi.com/1996-1944/8/8/5240/pdf&usg=AFQjCNFqudWbItH66ZYJLt6UvFMJZZ_WyQ&bvm=bv.102022582,d.d24
.....
Althought the phrase
" Although the use of artificial light such as torch light to trigger an early
blossom had been reported in ancient China nearly a thousand years ago ..."
Makes you wonder ....



* photosynthetic action spectrum resemblance (SRPAS)
photosynthetic action spectrum (PAS)

( ...) It is noteworthy that plants do absorb green light to some significant
extent, e.g., the absorption of green emissions at 555 nm, for example, is 26% of the peak absorption in
the PAS. Furthermore, the energy absorbed in the green light region, i.e., from 495 to 570 nm, measures
17% of the total energy absorbed by the photosynthetic action spectrum. This implies that the green
light-dominant mid-wavelength emission is not to be ignored in plant growth [48].


The high spectral resemblance may be attributed to the employment of a twin-peak blue emitter
thatnerates two broad-bands covering the short- to mid-wavelength regions, and the employment of a
diffused mono-peak red emitter that generates a relatively wide broad-band extending from the mid- to
long-wavelength regions. In addition, an over-90% spectrum resemblance can also be obtainable,
provided a deeper red emitter is incorporated, as shown in Figure 3b.
It is interesting to find that the typical LED lamps (Figure 4a) show a SRPAS higher than that of the
plant growth-specific LED. That is because the light sources of the former emit a broad band of light
ranging from 470 to at least 780 nm and, hence, a much wider overlap with the PAS results in the
mid-wavelength region, although the overlap is somewhat lower in the red emission.
To improve on this, the inclusion of more red and blue emissions are suggested in typical white LED
lamps. For example, the SRPAS can be increased from 60% to 91% as two additional blue and red LEDs
peaking at the vicinity of the respective absorption peaks of the PAS are employed. (Figure 4b).
(....)


So ...
These folks seem to know pretty much their shit ,ehh?
..... they seem to know ,oh yeah ......
How come in China you did not .?
?
The 80's are over,mate ...
History .Gone ...
For good ...

.....
425nm x3
445nm x9
450nm x6
475nm x4
480nm x4
640nm x7
660nm x14
675nm x7
730n x2
6500k x2

12000k x2 (<==== !!!! Ain't that known also as "blue " ? )
>Oh my ....My Gawd ....
...
Wher's the dried frog leggiess and the worm-spit tincture ?
Without them ,yain't gonna be doin' anythin' right ,son ...

Better yet ,trust the Taiwanese ...
They seem to know ,much better ...

 

[unlmtd216]

Go buy 2 Cree cxb 3070's, 2 alpine 11 plus cpu coolers, and a 100 watt driver. The XPE chips can be anywhere from 60 to 120 lumen per watt. Most likely it will be closer to 60 lumen per watt. They aren't going to source high bin chips, unless you specify what you want. At these outputs you might as well use cfl's. The crappiest CXB 3070 bin @ 50 watts will get you over 110 lumen per watt minimum and best bin @ 50 watts is over 170 lumen per watt.

Hmm..that sounds like a huge difference in efficiency and I am interested in DIY LED's....How much do you think it would cost to build the system you described? and does it utilize the "Mcree Curve"? I really am only buying from a manufacturer because I'm looking to get into business....but perhaps I should become a manufacturer seeing as most of the existing companies seem to have little knowledge on the subject. I would like to gain as much knowledge as possible!!! and I appreciate it.

 

[unlmtd216]

I'm not sure if this helps ...
Maybe in China still the all-mighty "magic spectrum " remains unknown ...
but for sure ,in Taiwan ,they know plenty more ,than the chinese .......

Wher's the dried frog leggiess and the worm-spit tincture ?
Without them ,yain't gonna be doin' anythin' right ,son ...
Trust the Taiwanese ...
They know ...

Interesting post! I wish I knew what all those values and formulas meant...lol that just shows how bored I'd get studying this stuff...I need a simple formula that works...like percentages...seems I got a lot to learn before I can KNOW for myself what works BEST. I appreciate it nonetheless !

 

[OneHitDone]

I'm not sure if this helps ...
Maybe in China still the all-mighty "magic spectrum " remains unknown ...
but for sure ,in Taiwan ,they know plenty more ,than the chinese .......

https://www.google.gr/url?sa=t&rct=j&q=&esrc=s&source=web&cd=42&cad=rja&uact=8&ved=0CB0QFjABOChqFQoTCMey5sL14scCFQZtFAodeNwH-Q&url=http://www.mdpi.com/1996-1944/8/8/5240/pdf&usg=AFQjCNFqudWbItH66ZYJLt6UvFMJZZ_WyQ&bvm=bv.102022582,d.d24
.....
Althought the phrase
" Although the use of artificial light such as torch light to trigger an early
blossom had been reported in ancient China nearly a thousand years ago ..."
Makes you wonder ....



* photosynthetic action spectrum resemblance (SRPAS)
photosynthetic action spectrum (PAS)

( ...) It is noteworthy that plants do absorb green light to some significant
extent, e.g., the absorption of green emissions at 555 nm, for example, is 26% of the peak absorption in
the PAS. Furthermore, the energy absorbed in the green light region, i.e., from 495 to 570 nm, measures
17% of the total energy absorbed by the photosynthetic action spectrum. This implies that the green
light-dominant mid-wavelength emission is not to be ignored in plant growth [48].


The high spectral resemblance may be attributed to the employment of a twin-peak blue emitter
thatnerates two broad-bands covering the short- to mid-wavelength regions, and the employment of a
diffused mono-peak red emitter that generates a relatively wide broad-band extending from the mid- to
long-wavelength regions. In addition, an over-90% spectrum resemblance can also be obtainable,
provided a deeper red emitter is incorporated, as shown in Figure 3b.
It is interesting to find that the typical LED lamps (Figure 4a) show a SRPAS higher than that of the
plant growth-specific LED. That is because the light sources of the former emit a broad band of light
ranging from 470 to at least 780 nm and, hence, a much wider overlap with the PAS results in the
mid-wavelength region, although the overlap is somewhat lower in the red emission.
To improve on this, the inclusion of more red and blue emissions are suggested in typical white LED
lamps. For example, the SRPAS can be increased from 60% to 91% as two additional blue and red LEDs
peaking at the vicinity of the respective absorption peaks of the PAS are employed. (Figure 4b).
(....)


So ...
These folks seem to know pretty much their shit ,ehh?
..... they seem to know ,oh yeah ......
How come in China you did not .?
?
The 80's are over,mate ...
History .Gone ...
For good ...

.....
425nm x3
445nm x9
450nm x6
475nm x4
480nm x4
640nm x7
660nm x14
675nm x7
730n x2
6500k x2

12000k x2 (<==== !!!! Ain't that known also as "blue " ? )
>Oh my ....My Gawd ....
...
Wher's the dried frog leggiess and the worm-spit tincture ?
Without them ,yain't gonna be doin' anythin' right ,son ...

Better yet ,trust the Taiwanese ...
They seem to know ,much better ...

So would you say HPS defy's science? It sure seams to deliver