Cree CXA analysis

[PICOGRAV]

Quiz : What is the color of your shadow ,during a bright summer noun ?
Is it black ? Is it gray ? Is it dark brown ?
Or maybe dark blue ?

Place a dark blue paper sheet/object of some sort ,side by side with your shadow on a white/bright floor ....
(Next summer ,though ... )
Well ? What color is the shadow ?
( The opposite of the light creating it ..."Amber "/Warm sunlight creates blue shadows ...
Green light creates dark red shadows ...And so on ...
But the human brain always is "auto white-balancing " what we see ...
An A4 page will appear white ,under almost any kind of "white" light source ..
No matter if cool,neutral or warm ....
To notice the "real" color of an object (what it actually reflects back from any illuminating light source ),it
should be directly compared with something else of same( or different) known colour .

Have your fun now, I have explored more shadows then would expect to find on All Hallows' Evening

 

[Abiqua]

Quiz : What is the color of your shadow ,during a bright summer noun ?
Is it black ? Is it gray ? Is it dark brown ?
Or maybe dark blue ?

Place a dark blue paper sheet/object of some sort ,side by side with your shadow on a white/bright floor ....
(Next summer ,though ... )
Well ? What color is the shadow ?
( The opposite of the light creating it ..."Amber "/Warm sunlight creates blue shadows ...
Green light creates dark red shadows ...And so on ...
But the human brain always is "auto white-balancing " what we see ...
An A4 page will appear white ,under almost any kind of "white" light source ..
No matter if cool,neutral or warm ....
To notice the "real" color of an object (what it actually reflects back from any illuminating light source ),it
should be directly compared with something else of same( or different) known colour .

This is why I want to see a competent machine beyond a LUX or PAR meter for intensity and content. What about color blind people, CRT, CCI and looking at shadows to determine their color are moot, plus they don't look outside the PAR or Visible light spectrums, rarely either.

We see about 7-15 colors, while the "normal" person see's about, 150. Although we can see better without "light" at night too. So all this color shit to the human eye, for some, its moot.

I hadn't brought that into this, because I wanted to see how some of the convo played out.

 

[stardustsailor]

How does a lack of light, i.e. a shadow, have a color?

Is that not a kin to asking how loud silence is?

Well said and reasonably questioned ...

Actually "shadow" does not mean necessarily total darkness ( total absence of light ) .
What actually happens in the example / quiz is that light is 'subtracted' from where the shadow is cast uppon ...
(Or "black" is added ..... )
The weird thing is that is not actually "black " ...
Where a shadow is cast ,Colors of surfaces "darken" but also appear to "fill in" with the opposite color of the color of light illuminating the scene ...

( Artists/Painters /Graphic designers know about that "trick " ...)

So If one stands besides a White background ,and is lit by a Warm white light ,his/her shadow is not black ...
Neither any tone of grey ....
Actually is dark blue ....

The opposite happens if in the same scenery ,the light changes to areally cool ( bluish ) ..
Shadow ain't black ,ain't grey ,but ...dark yellow/amber ..(thus 'brown/tan' )
Placing a brown/tan colored object ,besides the casted shadow ,will easily confirm that ...


how loud silence is?


How loud is a thunder, when nobody's there to hear it ?
Silence can be extremely loud ...It depends mainly on the .." audio sensoring system" used ....
Matter is in a constant movement state of "existence" ..
Nothing is still ...
Not even at the Absolute Zero ( -273 °C ) ....
So does the air that surround us ...
It moves all the time ....
We,humans ,are able to sense only a fraction of that movement ..
And we call it "audio" ..Noise,music,voice,whatever ...
The rest ...Is it trully silence ?
Or we call "silence" our limitations ?
Our inabilities to hear further than ?
Oh yeah !
Trust me on that one ..
Silence is (or at least can be ) pretty loud ....

 

[Rrog]

So many layers beyond our perception. Thanks for some great dialog

 

[Beefbisquit]

When discussing concepts like 'silence', we need to understand that we have to discuss them within the limits of our perception.

If we fall into absurdity with our dialogue it might sound something like this;


"Hey man, what are you watching?"

"I can't be sure, there are infinite levels of existence and I'm only equipped to perceive a minuscule portion of all stimuli."

"I just wanted to know what was on TV."


Not very useful for dialogue.....

 

[Beefbisquit]

It is not absence of light, but just a different refracted spectrum, maybe described best by spectral irradiance. Even pitch black to the human eye has a "spectrum" or frequency if you will.

If there wasn't less light in shadows, PAR readings would be identical in the shade and direct sunlight.

 

[stardustsailor]

Ok,ok...

Let us end that ...
I hope I helped with the shadow color thing ...
As for the silence ..
Ok..
Let us say ,that "yes ,because we can't hear beyond that ( discuss them within the limits of our perception. ) so it must be silent and not loud by anymeans ..
Although.... it kinda reminds me of something like "And yet it moves " ...And everybody was discussing within the limits of own personal perception...
But then also that was not very useful for dialogue..
Heads used to fall -off from the shoulders ,actually ...
Funny thing ....
The question was simple and straightforward:
How loud is silence ?
Well what we call silence ?
What we can not hear ?
Or What really is happening ?
Cause if by "silence" we mean the absence of sound..
Well ,this can not be done nowhere else ,only inside our minds ...
Which is pretty 'relative' ...(different minds ,you see ...)
So ,there is not such thing .."Silence" ...
Only inside our minds ...Where "sound" is being born and dies ..
A pretty relative term ...Taken so seriously ....



As for your example ....
" See for yourself,what's on the TV.That's why you got the vision for .." ...

...Would have been my answer ...
Not very polite but more straightforward than the lame :
"I can't be sure, there are infinite levels of existence and I'm only equipped to perceive a minuscule portion of all stimuli."


Anyway ...Mr.Flux ...
Check this thread If you feel like it :

https://www.rollitup.org/led-other-lighting/646614-white-led-thermal-analysis-radiometric.html





But Caution : Maths converting photometric efficiency into radiometric , apply only for single phosphor white leds
(for example blue die and yellow YAG phosphor ,to make COOL/NEUTRAL white )
Those maths converting photometric efficiency into radiometric do not apply
for multiple phosphor mix/blend white leds ( like the high CRI ones ) .
The numbers /results do not seem to be correct at such case .




.....And yet is loud .

 

[Beefbisquit]

There is definitely objective silence.

Just like you might see red as orange, or purple or neon green; but if it's light in the 620-630nm range, it's red and only red. Your subjective experience of another color is irrelevant, as it's only your perception and not the objective reality of what's happening.

If I say the word "Computer" and you hear "calculator", your perception doesn't affect what was actually said, just like anyone's perception of colours doesn't actually affect things in reality.

 

[stardustsailor]

There is definitely objective silence.

Just like you might see red as orange, or purple or neon green; but if it's light in the 620-630nm range, it's red and only red. Your subjective experience of another color is irrelevant, as it's only your perception and not the objective reality of what's happening.

If I say the word "Computer" and you hear "calculator", your perception doesn't affect what was actually said, just like anyone's perception of colours doesn't actually affect things in reality.

Well ,I can't argue to what you say ...
Still keep in mind (about perception) that still is a 'relative' term ...
Like in the example with shadows of what is thoroughly analysed on the pdf Guod linked..
Our mind,oftenly 'play games' with us ( and our perception ) .....
For me Silence is when I do not hear any urban sounds/noise ...
For you it can be when you do not hear anything ...
For another ,silence is ..the sound of nature ( You know ..Birds singing ,springs hishing ,trees whistling ...)..
Yet we all mean " the absence of noise " by saying/thinking silence ...
We invented the term ,because in our mind it can exist ...
So ok ....Silence is silence and it can not be loud ...

( Yeap ..You're right .
It is impossible to convince a traffic cop ,that it was not actually "red" when you've just passed the lights ,but "deep orange " ....)

 

[PetFlora]

Maybe this will help... colors have a unique frequency, and therefor a unique sound

 

[SupraSPL]

Nice work! Thank you mrflux and welcome to RIU. I have been trying to talk folks out of high CRI because the published graphs are misleading. Your data really laid that to rest. It makes sense that hi CRI = more/different phosphor and lower radiometric efficiency.

It is very difficult for us layfolk to calculate radiometric efficiency for white LEDs but it is critical data if we want to build the most efficient grow lamps. To get that number for white LEDs we need LER. Up to this point I have just been estimating based on KNNAs LER recommendations from way back (280 for cool white, 300 for neutral and 310 for warm) Looks like we were way off a smidge? How did you arrive at your LER numbers, analyzing the curves or from manufacturer data charts of some kind?

For those who don't know what LER is, imagine an (impossible) 100% efficient LED. LER is how many lumens/watt it would produce. So a 100% efficient green LED would produce 683 lumens/watt. A 630nm red would max out at 190. White LEDs have complex curves so their LER is a calculus problem and it varies for each tint.

 

[SupraSPL]

Up to this point I have been using 3000K Cree XTE R3 bins and 4000K Cree XML2 T5 bin. I run them at 700mA which cost about $1.50-$2.50/watt. After studying the COB style LEDs (Cree CXA and Bridgelux Vero) it seems like they are more suited for growing because they can achieve the same efficiency and assembling the lights would be much simpler. A Vero 18 running at 700mA cost less than $1/watt so it is no problem to run them soft (20 watts). The thermal resistance is very low on the Vero (.5 degrees increase per dissipation watt) so the junction temps can be kept reasonably low just like the XTE and XML. As far as I can determine, the CXA is .65 degrees/watt.

So big thumbs up for COB style particularly the Vero.

 

[SupraSPL]

UV is generally considered the culprit for genetic damage. Different plants have different defenses and repair mechanisms. Visible light doesn't have the kind of energy to bust up DNA, methinks

good point

 

[SupraSPL]

Yes it is a bummer they didn't include 2700k in the graph. The Luxeon ES graphs are very detailed but they are all relatively high cri so we can't use for a direct comparison. The difference between 2700 and 3000 is a slightly shorter peak in blue and a slightly longer wavelength peak on the red side, probably a slightly lower LER?

Either way, I agree the 3000K Vero looks like a winner and we can add deep reds that are ~40% efficient (Luxeon ES and Osram SSL)

 

[PetFlora]

Do these new heat reductions eliminate the need for cooling fans?

Up to this point I have been using 3000K Cree XTE R3 bins and 4000K Cree XML2 T5 bin. I run them at 700mA which cost about $1.50-$2.50/watt. After studying the COB style LEDs (Cree CXA and Bridgelux Vero) it seems like they are more suited for growing because they can achieve the same efficiency and assembling the lights would be much simpler. A Vero 18 running at 700mA cost less than $1/watt so it is no problem to run them soft (20 watts). The thermal resistance is very low on the Vero (.5 degrees increase per dissipation watt) so the junction temps can be kept reasonably low just like the XTE and XML. As far as I can determine, the CXA is .65 degrees/watt.

So big thumbs up for COB style particularly the Vero.

 

[MrFlux]

Nice work! Thank you mrflux and welcome to RIU.

Hi Supra, glad you like it Things have come full circle now. When starting to look into the efficiency of white LEDs I of course googled it and came across some of your posts from years back. That is actually how I learned about LER!

Up to this point I have just been estimating based on KNNAs LER recommendations from way back (280 for cool white, 300 for neutral and 310 for warm) Looks like we were way off a smidge?

Those estimates for the LER seem too low, which would give inflated efficiency numbers. I'm generally seeing a LER of around 340 for Cree XTE, XML2, XBD and CXA, regardless of color temperature. Only high-CRI and extreme low color temp have a lower LER. Perhaps it used to be lower for the older generation of LEDs when KNNA was still active, it's really too bad we cannot ask him anymore.

How did you arrive at your LER numbers, analyzing the curves or from manufacturer data charts of some kind?

Yes I use the spectrum plots from the spec sheets. Or sometimes the textfile with spectrum numbers that is buried deep inside a gigabyte sized zipfile. From there on it gets cubic interpolation for extra smoothness, is multiplied with the CIE 1931 luminosity curve and then numerically integrated with 1 nm resolution. These are the hoops to jump through to get get an answer to the basic question of "what goes in and what comes out".

For those who don't know what LER is, imagine an (impossible) 100% efficient LED. LER is how many lumens/watt it would produce. So a 100% efficient green LED would produce 683 lumens/watt. A 630nm red would max out at 190. White LEDs have complex curves so their LER is a calculus problem and it varies for each tint.

Yep. With the LER number, together with the lumens per watt figure that the manifacturer proudly advertises, it is then very simple to know the efficiency: Efficiency in % = 100*(lumens per Watt) / LER.

For example, the Cree XP-E red has up to 73.9 lumen per Watt. The LER is found to be 201. The efficiency is then an amazing 100*73.9 / 201 = 51%. (since this is a 620-630 nm emitter the LER is slightly higher then for a pure 630nm peak).

 

[MrFlux]

^^^^ replace 73.9 with 103. Cannot edit the post anymore ("the message you entered is too short...")

 

[SupraSPL]

Do these new heat reductions eliminate the need for cooling fans?

Cooling requirements for COB style remain about the same as XML XTE Luxeon etc

Hi Supra, glad you like it Things have come full circle now. When starting to look into the efficiency of white LEDs I of course googled it and came across some of your posts from years back. That is actually how I learned about LER!

Ironically that is the same way I found your post. I googled something and your post on RIU popped up first. Well thanks for carrying it forward!

I use the spectrum plots from the spec sheets. Or sometimes the textfile with spectrum numbers that is buried deep inside a gigabyte sized zipfile. From there on it gets cubic interpolation for extra smoothness, is multiplied with the CIE 1931 luminosity curve and then numerically integrated with 1 nm resolution

Good enough for me I updated all my spreadsheets with your LER numbers. Here are the results:

All are at 700mA approx 50C junction temp:

Cree XTE 3000K R3 bin: 31.86% ($1.50/watt)
Cree CXA1820 3000K 30.8% ($0.52/watt)
Vero13 3000K 29.3% ($0.39/watt)
Vero18 3000K 34.44% ($0.95/watt)
Vero29 3000K 37.6% ($1.92/watt)
Cree XML2 3000K T3 34.34% ($2.25/watt)
Cree XML2 3000K T5 40.58% ($3.25/watt) (These just showed up at Cutter)
Cree XML2 4000K T5 40.11% ($2.25/watt)

 

[skyled]

Those estimates for the LER seem too low, which would give inflated efficiency numbers. I'm generally seeing a LER of around 340 for Cree XTE, XML2, XBD and CXA, regardless of color temperature. Only high-CRI and extreme low color temp have a lower LER. Perhaps it used to be lower for the older generation of LEDs when KNNA was still active, it's really too bad we cannot ask him anymore.


Yes I use the spectrum plots from the spec sheets. Or sometimes the textfile with spectrum numbers that is buried deep inside a gigabyte sized zipfile. From there on it gets cubic interpolation for extra smoothness, is multiplied with the CIE 1931 luminosity curve and then numerically integrated with 1 nm resolution. These are the hoops to jump through to get get an answer to the basic question of "what goes in and what comes out".

Hi ,
I've calculated several LERs to
but with 5nm resolution, because I don't have the luminosity curve, or its equation
I've only the number every 5nm
So, i've found 320-325 lm/w for the most of the White Leds (instead of your 340)

How do you do to integrate your curve with 1nm resolution?
And could you give me your luminosity curve (or equation if you have it) ?

Thx

 

[MrFlux]

So, i've found 320-325 lm/w for the most of the White Leds (instead of your 340)

That's already pretty close. Here is the list of every emitter that I have on file so that you can compare:

blue = 430-480nm, red = 620-680nm

XML2 2600K
Power in        : 2.0 W
Luminous flux   : 250 lumen
Efficacy        : 125 lumen/W
LER             : 355 lumen/W
Radiometric eff.: 35.2%
Radiant flux    : 0.70 W
Photon flux     : 3.46 umol/s
Blue            : 7% power, 5% flux
Red             : 26% power, 28% flux


XML2 3700K
Power in        : 2.0 W
Luminous flux   : 296 lumen
Efficacy        : 148 lumen/W
LER             : 344 lumen/W
Radiometric eff.: 43.0%
Radiant flux    : 0.86 W
Photon flux     : 3.93 umol/s
Blue            : 23% power, 19% flux
Red             : 13% power, 16% flux


XML2 5000K
Power in        : 2.0 W
Luminous flux   : 341 lumen
Efficacy        : 170 lumen/W
LER             : 334 lumen/W
Radiometric eff.: 51.0%
Radiant flux    : 1.02 W
Photon flux     : 4.55 umol/s
Blue            : 28% power, 24% flux
Red             : 10% power, 12% flux


Vero 3000K, 97 CRI
Power in        : 9.35 W
Luminous flux   : 755 lumen
Efficacy        : 81 lumen/W
LER             : 287 lumen/W
Radiometric eff.: 28.1%
Radiant flux    : 2.63 W
Photon flux     : 13.08 umol/s
Blue            : 9% power, 7% flux
Red             : 30% power, 33% flux


Vero 3000K
Power in        : 9.35 W
Luminous flux   : 1120 lumen
Efficacy        : 120 lumen/W
LER             : 337 lumen/W
Radiometric eff.: 35.5%
Radiant flux    : 3.32 W
Photon flux     : 16.42 umol/s
Blue            : 8% power, 6% flux
Red             : 27% power, 30% flux


Vero 4000K
Power in        : 9.35 W
Luminous flux   : 1180 lumen
Efficacy        : 126 lumen/W
LER             : 341 lumen/W
Radiometric eff.: 37.0%
Radiant flux    : 3.46 W
Photon flux     : 16.65 umol/s
Blue            : 12% power, 10% flux
Red             : 23% power, 25% flux


Vero 5000K
Power in        : 9.35 W
Luminous flux   : 1305 lumen
Efficacy        : 140 lumen/W
LER             : 352 lumen/W
Radiometric eff.: 39.7%
Radiant flux    : 3.71 W
Photon flux     : 17.11 umol/s
Blue            : 21% power, 17% flux
Red             : 15% power, 17% flux


XBD 2600K
Power in        : 1.0 W
Luminous flux   : 114 lumen
Efficacy        : 114 lumen/W
LER             : 342 lumen/W
Radiometric eff.: 33.3%
Radiant flux    : 0.33 W
Photon flux     : 1.62 umol/s
Blue            : 10% power, 8% flux
Red             : 25% power, 27% flux


XBD 3700K
Power in        : 1.0 W
Luminous flux   : 130 lumen
Efficacy        : 130 lumen/W
LER             : 343 lumen/W
Radiometric eff.: 37.9%
Radiant flux    : 0.38 W
Photon flux     : 1.80 umol/s
Blue            : 16% power, 13% flux
Red             : 19% power, 22% flux


XBD 5000K
Power in        : 1.0 W
Luminous flux   : 139 lumen
Efficacy        : 139 lumen/W
LER             : 336 lumen/W
Radiometric eff.: 41.3%
Radiant flux    : 0.41 W
Photon flux     : 1.85 umol/s
Blue            : 24% power, 20% flux
Red             : 11% power, 13% flux


XTE 2200K
Power in        : 1.0 W
Luminous flux   : 123 lumen
Efficacy        : 123 lumen/W
LER             : 325 lumen/W
Radiometric eff.: 37.8%
Radiant flux    : 0.38 W
Photon flux     : 1.90 umol/s
Blue            : 6% power, 5% flux
Red             : 31% power, 33% flux


XTE 2600K 85-CRI
Power in        : 1.0 W
Luminous flux   : 100 lumen
Efficacy        : 100 lumen/W
LER             : 341 lumen/W
Radiometric eff.: 29.4%
Radiant flux    : 0.29 W
Photon flux     : 1.44 umol/s
Blue            : 9% power, 7% flux
Red             : 26% power, 29% flux


XTE 3700K
Power in        : 1.0 W
Luminous flux   : 130 lumen
Efficacy        : 130 lumen/W
LER             : 341 lumen/W
Radiometric eff.: 38.2%
Radiant flux    : 0.38 W
Photon flux     : 1.82 umol/s
Blue            : 16% power, 13% flux
Red             : 21% power, 24% flux


XTE 5000K
Power in        : 1.0 W
Luminous flux   : 160 lumen
Efficacy        : 160 lumen/W
LER             : 328 lumen/W
Radiometric eff.: 48.8%
Radiant flux    : 0.49 W
Photon flux     : 2.19 umol/s
Blue            : 29% power, 25% flux
Red             : 11% power, 13% flux


XPE red
Power in        : 0.72 W
Luminous flux   : 73.9 lumen
Efficacy        : 103 lumen/W
LER             : 201 lumen/W
Radiometric eff.: 51.0%
Radiant flux    : 0.37 W
Photon flux     : 1.94 umol/s
Blue            : 0% power, 0% flux
Red             : 88% power, 89% flux


XTE Royal Blue
Power in        : 1.0 W
Luminous flux   : 30.8 lumen
Efficacy        : 31 lumen/W
LER             : 51 lumen/W
Radiometric eff.: 60.0%
Radiant flux    : 0.60 W
Photon flux     : 2.27 umol/s
Blue            : 95% power, 95% flux
Red             : 0% power, 0% flux


CXA 3000K, 93 CRI
Power in        : 3.8 W
Luminous flux   : 317 lumen
Efficacy        : 83 lumen/W
LER             : 291 lumen/W
Radiometric eff.: 28.7%
Radiant flux    : 1.09 W
Photon flux     : 5.36 umol/s
Blue            : 12% power, 9% flux
Red             : 26% power, 29% flux


CXA 3000K
Power in        : 3.8 W
Luminous flux   : 423 lumen
Efficacy        : 111 lumen/W
LER             : 344 lumen/W
Radiometric eff.: 32.3%
Radiant flux    : 1.23 W
Photon flux     : 5.98 umol/s
Blue            : 10% power, 8% flux
Red             : 24% power, 27% flux


CXA 4000K
Power in        : 3.8 W
Luminous flux   : 490 lumen
Efficacy        : 129 lumen/W
LER             : 339 lumen/W
Radiometric eff.: 38.1%
Radiant flux    : 1.45 W
Photon flux     : 6.77 umol/s
Blue            : 19% power, 15% flux
Red             : 17% power, 20% flux


CXA 5000K
Power in        : 3.8 W
Luminous flux   : 527 lumen
Efficacy        : 139 lumen/W
LER             : 339 lumen/W
Radiometric eff.: 40.9%
Radiant flux    : 1.55 W
Photon flux     : 7.16 umol/s
Blue            : 23% power, 19% flux
Red             : 14% power, 17% flux


Oslon SSL 2700K
Power in        : 1.12 W
Luminous flux   : 82 lumen
Efficacy        : 73 lumen/W
LER             : 285 lumen/W
Radiometric eff.: 25.7%
Radiant flux    : 0.29 W
Photon flux     : 1.45 umol/s
Blue            : 6% power, 4% flux
Red             : 33% power, 36% flux

How do you do to integrate your curve with 1nm resolution?
And could you give me your luminosity curve (or equation if you have it) ?

The integration is just summing every factor, per 1 nm bin. The 1nm CIE curve is attached.

How do you digitize the spectrum? What I do is to sample the curve with a couple of dozen points (with tops and bottoms) and then use cubic interpolation. Using just linear interpolation would not work well.