Spectrum shift in COBS due to dimming/current reduction....

thetr33man

thetr33man

Well-Known Member
So I made my first light and snapped some pictures. 2 really caught my attention and Im wondering what everyone's opinion is on this. When I dimmed my light, its obvious the color temperature changes significantly. Compare the 2 photos below.... This light is made using cxb2530s 3000k 90cri, first picture is dimmed significantly, second is full power @ 700mA.



I did a little research and found this article....
http://www.lrc.rpi.edu/programs/solidstate/cr_effectsofdimming.asp

So basically what Im thinking is, if you want 3500k light, you really need to consider how much power you are going to be supplying to the cob. 36v CxB3590's are rated at 2.4A, so if you buy a 3500k cob rated at this power, its color is going to shift significantly towards the red end of the spectrum if it is run at 1400mA and even more at 700mA. Something to consider... I would be really interested to see if cree has run any tests showing this color shift in a graph form. Maybe if you want your plants to get 3500k light but you are only running 1A, you would need to get 5000k cobs instead....
 
sixstring2112

sixstring2112

Well-Known Member
caretak3r said:
wow, that's a great observation and kind of a can of worms for the low amperage cxb crowd :)
Click to expand...
yeah pretty much for everyone because i dont see many people using them at full power.makes me want to drive my 3000k harder so they dont drift down below the hps spectrums of 2000 to 2700k.quite the can you opened @thetr33man lol.
 
thetr33man

thetr33man

Well-Known Member
It may have been covered, if anyone has more info Id like to see it. Its obvious the spectrum shifts to the red, just the amount Im not sure of.
 
sanjuan

sanjuan

Well-Known Member
Where might Mr Flux material be located? (There is a MR FLuX registered here with no posts.)
 
Scotch089

Scotch089

Well-Known Member
I may be wrong, but I feel like it was flux, sds, and guod (I hope you are doing well sir!!) Talking about it, not profound... less blue light passing through a yellow phosphor.... but I remember someone actually laying out graphs and data

Ringing any bells guod?
 
BOBBY_G

BOBBY_G

Well-Known Member
caretak3r said:
wow, that's a great observation and kind of a can of worms for the low amperage cxb crowd :)
Click to expand...
maybe this is why @Growmau5 crushes it so hard with his 700 ma 3500k rig - its really running 3000k

of note this is great for the fact that higher color temps bin higher.

and as you dim them they get more efficient

in other words some 4000k db bin when dimmed may out perform the same 3500k cd bins (which youd need to push 8% harder to get the same brightness)

ill have to wrap my head around that one for awhile
 
superbak3d

superbak3d

Well-Known Member
Correct me if I am wrong.

But sure if you dim down the color temp gets warmer, but at the same time aren't you killing your PPFD? Wouldn't you need to basically double up on the number of cobs you're running since they're driven so low?
 
testiclees

testiclees

Well-Known Member
Scotch089 said:
@caretak3r give this a read, starts talking about 670 at page 2, sold me on it..

https://www.rollitup.org/t/photosynthesis-under-solid-state-light-setting-the-standards.833449/page-2
Click to expand...
Post 21 (in that thread) discusses light periods in relation to yield and metabolites. Do we know if those plants went to lights out with actual sunset or if it was a blackout cloth? Article states that the experiment is conducted in a greenhouse AND that they practice a structured photoperiod. Im interested because the results might be very different if the plants were exposed to 730nm at end of day.
 
FranJan

FranJan

Well-Known Member
http://www.digikey.com/en/articles/techzone/2012/sep/digital-dimming-solves-led-color-dilemma
"This article considers why changes in voltage affect the light output and color emitted by an LED. The article then describes analog and digital dimming techniques before concluding which is best for mitigating the worst effects on light quality caused by current regulation

http://www.digikey.com/en/articles/techzone/2012/jun/led-dimming-solutions
"This article has discussed the challenges and solutions of LED dimming. It has reviewed the two main categories of LED dimming: PWM and analog, noting that both have advantages and disadvantages: Analog dimming is usually a simpler circuit, but may be inappropriate for applications that require a constant color temperature; PWM dimming does not have the same issues with color changes in the LED for varying brightness levels, but requires additional logic to create the PWM waveforms. For more information on the products mentioned here, use the links provided to access product information pages on the Digi-Key website."

Somewhere out there there's a small company that makes a LED driver (patented) that does perfect analog dimming, can't remember. @guod???? Maybe someone else too, even MW?

I've oftened wondered have any LED growers stepped in a bucket of shit and ended up changing their LED's spectral output to something their plants really really liked and got something extra from it? Personally to me a COB is just a photon pusher, I could give a you-know-what about it's GENERAL color as long as it's in the ballpark of maintaining its radiated output. You want color OSRAM/Philips/CREE are always open for the business :P but yeah I would think with COB only grows you want to maintain your color consistency.


 
MrTwist1

MrTwist1

Well-Known Member
thetr33man said:
So I made my first light and snapped some pictures. 2 really caught my attention and Im wondering what everyone's opinion is on this. When I dimmed my light, its obvious the color temperature changes significantly. Compare the 2 photos below.... This light is made using cxb2530s 3000k 90cri, first picture is dimmed significantly, second is full power @ 700mA.



I did a little research and found this article....
http://www.lrc.rpi.edu/programs/solidstate/cr_effectsofdimming.asp

So basically what Im thinking is, if you want 3500k light, you really need to consider how much power you are going to be supplying to the cob. 36v CxB3590's are rated at 2.4A, so if you buy a 3500k cob rated at this power, its color is going to shift significantly towards the red end of the spectrum if it is run at 1400mA and even more at 700mA. Something to consider... I would be really interested to see if cree has run any tests showing this color shift in a graph form. Maybe if you want your plants to get 3500k light but you are only running 1A, you would need to get 5000k cobs instead....
Click to expand...
Hey those pictures don't mean too much unless the camera was set to a custom white balance. If the camera was (as most are) set to "auto white balance" then the pictures really show nothing useful since the camera itself is making an interpretation of the colour of the scene. If you look carefully you will see the second picture has a much bluer tint to it in general - and not only where the light of the COBs strike. Look at the lighter areas at the back of the images and you can see the second image is much bluer here. It seems almost certain that these images were shot with "Auto White Balance".

It seems likely that there is some small colour temp shift according to how hard COBs are driven, but not so dramatically as these images suggest. Cheers
 
You must log in or register to reply here.
Top