The difference between relative and real spectral power

l0wbob2016

l0wbob2016

Well-Known Member
Hello folks,

it passed some time since i was active in this forum, but this doesnt mean that i didnt learn anything new.

Something i've come along is the difference between the relative spectral power distribution for LED's / COB's that you can find in their Datasheets and the real spectral power distribution given at a certain power.

At first when i saw comparisions between the relative and the real spectral power distribution i said this cant be true. So i said to myself, i need to learn this stuff and check if it is correct.

The final moment of truth after double-checking 4-5 times came and i am pretty sure i got it correct. ( If not pls correct me )

My personal conclusion of this was, that comparing different color-spectra from LED's / COB's with the given spectrum from the datasheet is the wrong way in terms of light for plants. Dont get me wrong, the relative spectral power distribution charts have their meaning and right to exist, but not in terms of comparing color spectra for plants.

To show you the difference i've made a comparison-chart. I compared the Bridgelux Vero29-C Versions ( 69,4V ) only.
The versions i compared where the following:
BXRC-27G10K0-C-7X
BXRC-30E10K0-C-73
BXRC-40E10K0-C-7X
BXRC-56G10K0-C-74
BXRC-65C10K1-C-7x

and here is the chart:



The dashed lines are the relatives from the datasheet and the normal lines are the real spectral power distribution ( calculated by myself ) at a power of 56,6W.

As you can see, calculating the real spectral power distribution gives a much better overview and is the only way to compare different LED's / COB's in terms of light for plants.

For comparing it for plants, there are some more steps to go, but for all steps you need the real spectral power distribution.
To go more into detail for all tested COB's here are some values i calculated:

BXRC-27G10K0-C-7X CRI90
56,6 W | 8610 lm | 152,12 lm/W | lightpower 32,08W | efficiency 56,69% | PAR PPF 148,87 µmol/s , 2,63 µmol/J , 17,29 PAR/klx | radiation angle 120° distance 50cm ~ 10900 lx |
calc. max PPFD centerpoint 189,55 µmol/m²s

BXRC-30E10K0-C-73 CRI80
56,6 W | 10763 lm | 190,16 lm/W | lightpower 32,32W | efficiency 57,09% | PAR PPF 151,65 µmol/s , 2,68 µmol/J , 14,09 PAR/klx| radiation angle 120° distance 50cm ~ 13700 lx |
calc. max PPFD centerpoint 193,08 µmol/m²s

BXRC-40E10K0-C-7X CRI80
56,6 W | 11179 lm | 197,51 lm/W | lightpower 33,73W | efficiency 59,59% | PAR PPF 155,15 µmol/s , 2,74 µmol/J , 13,88 PAR/klx| radiation angle 120° distance 50cm ~ 14200 lx |
calc. max PPFD centerpoint 197,54 µmol/m²s

BXRC-56G10K0-C-74 CRI90
56,6 W | 10277 lm | 181,57 lm/W | lightpower 36,98W | efficiency 65,34% | PAR PPF 164,61 µmol/s , 2,91 µmol/J , 16,02 PAR/klx| radiation angle 120° distance 50cm ~ 13000 lx |
calc. max PPFD centerpoint 209,58 µmol/m²s

BXRC-65C10K1-C-7x CRI70
56,6 W | 12082 lm | 213,46 lm/W | lightpower 36,06W | efficiency 63,70% | PAR PPF 159,94 µmol/s , 2,82 µmol/J , 13,24 PAR/klx| radiation angle 120° distance 50cm ~ 15300 lx |
calc. max PPFD centerpoint 203,65 µmol/m²s

Any thoughts/ideas/input about this?
 
CobKits

CobKits

Well-Known Member
light quantity and quality are two different things.

once you decide on the spectrum you need, light quantity can be quantified and adjusted.

unless you are really looking to photoinhibit with blue its not super critical which is why 2700k-4000k isnt all that different to plants

if you are trying to match the same par levels with a 6500k as a 2700k the former will be getting absolutely pounded with blue which will have a noticeable effect
 
l0wbob2016

l0wbob2016

Well-Known Member
CobKits said:
light quantity and quality are two different things.

once you decide on the spectrum you need, light quantity can be quantified and adjusted.

unless you are really looking to photoinhibit with blue its not super critical which is why 2700k-4000k isnt all that different to plants

if you are trying to match the same par levels with a 6500k as a 2700k the former will be getting absolutely pounded with blue which will have a noticeable effect
Click to expand...
I find it just very interesting that comparing the real spectral power to the normalized one in the datasheet.
The COB's in my chart are all at the same power. The "green" of a 6500K is as high as the "red" of the 3000K. Thats more than twice the amount of "green" the 6500K produces.
If i would mix a 6500K and a 3000K with the same wattage, the colorshift would be much more than i would expect when comparing the color spectrum from the datasheets.
 
alesh

alesh

Well-Known Member
l0wbob2016 said:
Hello folks,

it passed some time since i was active in this forum, but this doesnt mean that i didnt learn anything new.

Something i've come along is the difference between the relative spectral power distribution for LED's / COB's that you can find in their Datasheets and the real spectral power distribution given at a certain power.

At first when i saw comparisions between the relative and the real spectral power distribution i said this cant be true. So i said to myself, i need to learn this stuff and check if it is correct.

The final moment of truth after double-checking 4-5 times came and i am pretty sure i got it correct. ( If not pls correct me )

My personal conclusion of this was, that comparing different color-spectra from LED's / COB's with the given spectrum from the datasheet is the wrong way in terms of light for plants. Dont get me wrong, the relative spectral power distribution charts have their meaning and right to exist, but not in terms of comparing color spectra for plants.

To show you the difference i've made a comparison-chart. I compared the Bridgelux Vero29-C Versions ( 69,4V ) only.
The versions i compared where the following:
BXRC-27G10K0-C-7X
BXRC-30E10K0-C-73
BXRC-40E10K0-C-7X
BXRC-56G10K0-C-74
BXRC-65C10K1-C-7x

and here is the chart:



The dashed lines are the relatives from the datasheet and the normal lines are the real spectral power distribution ( calculated by myself ) at a power of 56,6W.

As you can see, calculating the real spectral power distribution gives a much better overview and is the only way to compare different LED's / COB's in terms of light for plants.

For comparing it for plants, there are some more steps to go, but for all steps you need the real spectral power distribution.
To go more into detail for all tested COB's here are some values i calculated:

BXRC-27G10K0-C-7X CRI90
56,6 W | 8610 lm | 152,12 lm/W | lightpower 32,08W | efficiency 56,69% | PAR PPF 148,87 µmol/s , 2,63 µmol/J , 17,29 PAR/klx | radiation angle 120° distance 50cm ~ 10900 lx |
calc. max PPFD centerpoint 189,55 µmol/m²s

BXRC-30E10K0-C-73 CRI80
56,6 W | 10763 lm | 190,16 lm/W | lightpower 32,32W | efficiency 57,09% | PAR PPF 151,65 µmol/s , 2,68 µmol/J , 14,09 PAR/klx| radiation angle 120° distance 50cm ~ 13700 lx |
calc. max PPFD centerpoint 193,08 µmol/m²s

BXRC-40E10K0-C-7X CRI80
56,6 W | 11179 lm | 197,51 lm/W | lightpower 33,73W | efficiency 59,59% | PAR PPF 155,15 µmol/s , 2,74 µmol/J , 13,88 PAR/klx| radiation angle 120° distance 50cm ~ 14200 lx |
calc. max PPFD centerpoint 197,54 µmol/m²s

BXRC-56G10K0-C-74 CRI90
56,6 W | 10277 lm | 181,57 lm/W | lightpower 36,98W | efficiency 65,34% | PAR PPF 164,61 µmol/s , 2,91 µmol/J , 16,02 PAR/klx| radiation angle 120° distance 50cm ~ 13000 lx |
calc. max PPFD centerpoint 209,58 µmol/m²s

BXRC-65C10K1-C-7x CRI70
56,6 W | 12082 lm | 213,46 lm/W | lightpower 36,06W | efficiency 63,70% | PAR PPF 159,94 µmol/s , 2,82 µmol/J , 13,24 PAR/klx| radiation angle 120° distance 50cm ~ 15300 lx |
calc. max PPFD centerpoint 203,65 µmol/m²s

Any thoughts/ideas/input about this?
Click to expand...
Yes, relative SPDs normalized to 1 could be quite misleading for an untrained eye. Shouldn't be news to anyone there but I guess it is. I'd rather call the adjusted SPD absolute than real.

How did you calulate the centerpoint illuminance and photon density? This is my pedantic me speaking but there should be either PPF/klm or PPFD/klx.
 
l0wbob2016

l0wbob2016

Well-Known Member
alesh said:
Yes, relative SPDs normalized to 1 could be quite misleading for an untrained eye. Shouldn't be news to anyone there but I guess it is. I'd rather call the adjusted SPD absolute than real.

How did you calulate the centerpoint illuminance and photon density? This is my pedantic me speaking but there should be either PPF/klm or PPFD/klx.
Click to expand...
hehe, i also thought about calling it absolute, but since you have to calculate it for every power again, i called it real SPD for that power... maybe just a personal thing.

yes you are right, that was a typo, gonna change it to PPF/klm. ( well ok i cant change it anymore )

The calculation i've done with these steps:
- get a digitizer and digitize the relative SPD for the specific COB
- transform the relative SPD to (real) absolute SPD
- radiation angle of 120° and distance to meassurepoint 50cm
- everything done with excel

PAR/area
 
alesh

alesh

Well-Known Member
l0wbob2016 said:
hehe, i also thought about calling it absolute, but since you have to calculate it for every power again, i called it real SPD for that power... maybe just a personal thing.

yes you are right, that was a typo, gonna change it to PPF/klm. ( well ok i cant change it anymore )

The calculation i've done with these steps:
- get a digitizer and digitize the relative SPD for the specific COB
- transform the relative SPD to (real) absolute SPD
- radiation angle of 120° and distance to meassurepoint 50cm
- everything done with excel

PAR/area
Click to expand...
Could you be more specific about that centerpoint calculation? I think you'd also need to digitize spatial distribution pattern. Then calculate how big part of emitted light is being radiated in a small angle (5° seems reasonable) in center downwards position. Then you could estimate average illuminance in a circle projected by the angle (about 5 cm diameter at 50 cm height). Not sure how accurate it'd be. Not much I'd say.
 
l0wbob2016

l0wbob2016

Well-Known Member
alesh said:
Could you be more specific about that centerpoint calculation? I think you'd also need to digitize spatial distribution pattern. Then calculate how big part of emitted light is being radiated in a small angle (5° seems reasonable) in center downwards position. Then you could estimate average illuminance in a circle projected by the angle (about 5 cm diameter at 50 cm height). Not sure how accurate it'd be. Not much I'd say.
Click to expand...
A LED / COB is almost a Lambertian emitter.
Taking this link: https://en.wikipedia.org/wiki/Lambert's_cosine_law ( keep steradiant in mind )
and assuming that cos(θ). [ with θ=90° ]

So my value calculated is the theoretical maximum in the perfect 90° angle under the light source
 
Evil-Mobo

Evil-Mobo

Well-Known Member
l0wbob2016 said:
I find it just very interesting that comparing the real spectral power to the normalized one in the datasheet.
The COB's in my chart are all at the same power. The "green" of a 6500K is as high as the "red" of the 3000K. Thats more than twice the amount of "green" the 6500K produces.
If i would mix a 6500K and a 3000K with the same wattage, the colorshift would be much more than i would expect when comparing the color spectrum from the datasheets.
Click to expand...
I just mixed my 6500K/3000K cobs in one of my spaces and the plants are happy as can be.........
 
Last edited:
tstick

tstick

Well-Known Member
I've said it before and I'll say it again...

Get a cat.

Put it in a room with several different types of grow lights.

Close the door.

Wait for 10-15 minutes or so.

Come back in and observe which light the cat is sitting under.

Whichever light it is will be the one that your plants will like the best.

FACT.

;)
 
tstick

tstick

Well-Known Member
Also...just FYI...

Cats can also be used for phenotype hunting.

Put the cat into your grow area with all your phenotypes.

Close the door.

Wait 10-15 minutes or so.

Come back in and observe which plant the cat has chewed to pieces.

Whichever plant that is, is your best phenotype.

FACT.

;)
 
GroErr

GroErr

Well-Known Member
tstick said:
Also...just FYI...

Cats can also be used for phenotype hunting.

Put the cat into your grow area with all your phenotypes.

Close the door.

Wait 10-15 minutes or so.

Come back in and observe which plant the cat has chewed to pieces.

Whichever plant that is, is your best phenotype.

FACT.

;)
Click to expand...
^^^ fact ^^^

My cat would only chew the leaves from the dankest phenos if I wasn't paying attention. Of course leaving them in the room for any length of time wouldn't be the best way to preserve the best phenos - lol
 
hybridway2

hybridway2

Amare Shill
tstick said:
I've said it before and I'll say it again...

Get a cat.

Put it in a room with several different types of grow lights.

Close the door.

Wait for 10-15 minutes or so.

Come back in and observe which light the cat is sitting under.

Whichever light it is will be the one that your plants will like the best.

FACT.

;)
Click to expand...
I bet you're right.
But they steal your soul when you're sleeping.
Good to see you back, how you been 10wbob?
 
You must log in or register to reply here.
Top