G6 Blue spectrum mystery

[Jon's Reef]

I have been looking at the G5 to G6 blue changes and the spectrum is a bit puzzling.

G5 XR30 Blue 430/Royal Blue/ Blue diode count: 6/40/24
G6 XR30 Blue 430/Royal Blue/ Blue diode count: 12/32/32

This is the data published by Ecotech and similar to results from BRS testing:

Doing a basic spectral reconstruction of the G5, then using the diode ratios to adjust, we should expect something like this:
(200% of 430, 80% of the Royal (450 nm), 33% more Blue (~460 nm))

Looking at the BRS Pro spectra, it almost looks like the 400-500 nm spectrum is identical to the blue spectra even though the diode ratio is significantly different.
G5 XR30 Pro 430/Royal Blue/ Blue diode count: 4/32/24
G6 XR30 Pro 430/Royal Blue/ Blue diode count: 8/32/16

Note that the Pro spectra does seem to be inline with the diode ratio adjustments.
(double 430, same 450, 33% less 460)

Software bug? Driver limit? Power Supply limit?

If it was just the Ecotech data, it could be a graphic error on the website, but with BRS showing their graphs it appears to be real.

Very odd.....

 

[oreo54]

I have been looking at the G5 to G6 blue changes and the spectrum is a bit puzzling.

G5 XR30 Blue 430/Royal Blue/ Blue diode count: 6/40/24
G6 XR30 Blue 430/Royal Blue/ Blue diode count: 12/32/32
This is the data published by Ecotech and similar to results from BRS testing:

Doing a basic spectral reconstruction of the G5, then using the diode ratios to adjust, we should expect something like this:
(200% of 430, 80% of the Royal (450 nm), 33% more Blue (~460 nm))

Looking at the BRS Pro spectra, it almost looks like the 400-500 nm spectrum is identical to the blue spectra even though the diode ratio is significantly different.
G5 XR30 Pro 430/Royal Blue/ Blue diode count: 4/32/24
G6 XR30 Pro 430/Royal Blue/ Blue diode count: 8/32/16

Note that the Pro spectra does seem to be inline with the diode ratio adjustments.
(double 430, same 450, 33% less 460)

Software bug? Driver limit? Power Supply limit?

If it was just the Ecotech data, it could be a graphic error on the website, but with BRS showing their graphs it appears to be real.

Very odd.....

Depends on the manuf and "exact" wavelengths of the diodes as well as the channel drive currents.
Another version using generic diodes.

G6 XR30 Blue 430/Royal Blue/ Blue diode count: 12/32/32

Doesn't brs adjust spectrums to "common useage" at times?

8:32:16

 

[Jon's Reef]

Depends on the manuf and "exact" wavelengths of the diodes as well as the channel drive currents.
Another version using generic diodes.

Doesn't brs adjust spectrums to "common useage" at times?

8:32:16

After seeing your post I started looking a bit further and found this article at the 1023 site:

Radion G6のスペクトル予想と検証→答え合わせ

ちょっと大阪に行ってたので遅れました。 EcotechMarineが公式にRadionの紹介ページを更新したので、前回の記事の答え合わせをしましょう。 まず、LED素子構成ですが、ほぼ予想通りでした。厳密には、395nm×2/405nm×4

1023world.net

Red is their prediction of the G6 Blue spectrum vs published

 

[oreo54]

Yea 440nm vs 450nm.....
Thing is the diodes they " seem" to use are 445.
All diodes come out in a range of electrical and spectral characteristics.
Not unusual to have 4 or more " bins" for a standard luxeon rebel royal blue.

At some point it gets to splitting hairs though.

 

[Jon's Reef]

More than just the 440 to 450, why is the 460/470 "Blue" so low.

It almost seems like the measurements were collected using a "G6 Pro" profile on the "G6 Blue"

Here is a blow up of the BRS measurements. Ignore the white lines (G5 Blue and Pro curves). The red numbers are the # of diodes. Isn't it odd how both have 32 Royals.... Blue has 32 blue and Pro 16 blue.... yet the Royal to Blue ratio is almost identical.

 

[sneekapeek]

looking at the lux of each units spectrum, seems a little low for wattage used.

 

[oreo54]

looking at the lux of each units spectrum, seems a little low for wattage used.

LUX doesn't really measure the whole spectrum accurately.
Based on lumens (lumens/meter or some such area = lux) which is measured in this range (photopic range).

1000 lux of 450nm blue light = 115.6 par
1000 lux of "daylight" = 23 PAR

 

[Jon's Reef]

It is interesting that the pic is so low on the blue compared to pro. It could be they tested an XR15 for one/ XR30 for other. Or it could be that the blue is limited by hardware/ software.

Also odd that BRS has delayed their full G6 content. 2-3 weeks ago they said in the comments of their G6 overview that they were going to release more thorough par and spectrum data.

 

[sneekapeek]

looking at the lux of each units spectrum, seems a little low for wattage used

LUX doesn't really measure the whole spectrum accurately.
Based on lumens (lumens/meter or some such area = lux) which is measured in this range (photopic range).

1000 lux of 450nm blue light = 115.6 par
1000 lux of "daylight" = 23 PAR

you see what I mean though, unit puts out 5700 lux at 200 watts, I know some units that put out twice the amount lux at a 4th of the watts. Or the same lux at 4th of watts. Something doesn’t add up! Par not that good on these units on a 30” deep tank. Makes me wonder efficiency wise why they do what they do!

 

[oreo54]

looking at the lux of each units spectrum, seems a little low for wattage used

you see what I mean though, unit puts out 5700 lux at 200 watts, I know some units that put out twice the amount lux at a 4th of the watts. Or the same lux at 4th of watts. Something doesn’t add up! Par not that good on these units on a 30” deep tank. Makes me wonder efficiency wise why they do what they do!

Depends on the amount of blue and the wavelength.
The more "white" diodes (and green) the higher the lumens but possibly the less par..
Since par is photon numbers it is also a measure of their efficiency.. i.e photons/watt which is more important than lumens/watt and the 2 are only partially related..
Like I demonstrated.. 450 nm blue at equal lux measurement has 5x the PAR..

That is why diodes like 450 are measured in mW not lumens. It makes no sense exactly or is confusing.

It would take 5000 lux of white leds to equal the "efficiency" (number of photons) of 1000 lux of blue leds
because lux meters only record (sample)
a fraction of the blue photons.

Simple lux meters use a green filter to "shape" the sensor sensitivity. This filter blocks red/blue photons.

not saying you are wrong but you need to look at the whole picture.

 

[sneekapeek]

Depends on the amount of blue and the wavelength.
The more "white" diodes (and green) the higher the lumens but possibly the less par..
Since par is photon numbers it is also a measure of their efficiency.. i.e photons/watt which is more important than lumens/watt and the 2 are only partially related..
Like I demonstrated.. 450 nm blue at equal lux measurement has 5x the PAR..

That is why diodes like 450 are measured in mW not lumens. It makes no sense exactly or is confusing.

It would take 5000 lux of white leds to equal the "efficiency" (number of photons) of 1000 lux of blue leds
because lux meters only record (sample)
a fraction of the blue photons.

Simple lux meters use a green filter to "shape" the sensor sensitivity. This filter blocks red/blue photons.

not saying you are wrong but you need to look at the whole picture.

So let’s say I have a fixture that uses .08 watts per 1 par, How do these fixtures compare? Is something lacking in their spectrum?

 

[damsels are not mean]

So let’s say I have a fixture that uses .08 watts per 1 par, How do these fixtures compare? Is something lacking in their spectrum?

PAR measurements (which are in the unit PPFD not PAR) have nothing to do with spectrum, unless your PAR meter is not sensitive enough to a particular wavelength (most of the ones we use in the hobby are visible-light only afaik). PPFD is the density of photons at a given point.

 

[sneekapeek]

PAR measurements (which are in the unit PPFD not PAR) have nothing to do with spectrum, unless your PAR meter is not sensitive enough to a particular wavelength (most of the ones we use in the hobby are visible-light only afaik). PPFD is the density of photons at a given point.

Then how would you compare efficiency when manufacturer doesn’t produce the specs for given fixture? Just trying to get a better understanding. Seems like it doesn’t matter and we are just spinning the wheel for fun!

 

[damsels are not mean]

Then how would you compare efficiency when manufacturer doesn’t produce the specs for given fixture? Just trying to get a better understanding. Seems like it doesn’t matter and we are just spinning the wheel for fun!

As far as efficiency it would come down to PAR measurements at a given wattage. That has nothing to do with spectrum, though spectrum may affect how efficiently it grows tings.

 

[Markreed84]

Peak Par will be lower because it's going through 3 layers of lens and being spread over a large area. That's how I see it anyway.

 

[Jon's Reef]

The original question though is not about PAR, rather about spectrum. If the number of Blue LEDs increased and the number of Royal Blue decreased in the G6, why does the spectrum not reflect this?

 

[Markreed84]

The original question though is not about PAR, rather about spectrum. If the number of Blue LEDs increased and the number of Royal Blue decreased in the G6, why does the spectrum not reflect this?

I can't answer this however if you watch the reef builders breakdown the LEDs in the G6 are totally different LEDs than the G5, they're a different size for example. So maybe the answer is somewhere in there

 

[Markreed84]

BTW I reckon the delay in a full review video is probs more to do with no randy anymore. Videos are slow these days

 

[oreo54]

Then how would you compare efficiency when manufacturer doesn’t produce the specs for given fixture? Just trying to get a better understanding. Seems like it doesn’t matter and we are just spinning the wheel for fun!

Well you are correct..but no lights have "par efficiency" or even real efficiency measurements.
Only o/t point was lumens isn't it..
In horticulture they use this:

PPF/W Typical (µmol/J)

See royal blue as an example:

LUXEON SunPlus 20 Line

LUXEON SunPlus 20 Line These high-performance LEDs incorporate a single focal height and a variety of colors, enabling customized color-tunable solutions. The LUXEON SunPlus 20 Line provides total spectrum control to optimize light output for each growth stage, maximizing crop yields and energy...

lumileds.com

Unfortunately that measurements only include 400-700nm
Newest "thinking" is to expand that to like 385-760nm
Short of a spectrometer the "best" considering the "new" thinking..(black hatch line)

Hard enough to get people not to think in lumens much less micromoles/joule.
Best way (full of errors) is "par"/watts.
Of course all at the same distance and using the same sensor and same environment (preferably "in-tank")
And a random sample off the assembly line and not some cherry picked unit.

Like everything it's complicated..even for something simple like white leds (big royal blue under-measured peak)

Seems like it doesn’t matter and we are just spinning the wheel for fun!

Answer is yes and no.. Want to complicate it more?
Delivery.. you can have a low efficiency diode give great par numbers/watt IF it is more closely focused.
Say 90 vs 120 degree lenses.
Or different "bins" of a diode produces different efficiencies. A CREE xt-e royal blue from bin x,y,or z will have different efficiencies so even knowing "the diode model" doesn't help.

Testing lights using an integrating sphere and measuring in the expanded par range would be "ideal"..or at least order of magnitude better.
I'm sure Phillips did it.. and "maybe" Kessil. If they did neither published numbers for consumers.

Anyways all just food for thought really..