Just keep in mind due to the slightly different blue bases the Epileds spectra will not be the same as the Epistar based chips. The phosphor is the same so the 20W, 50W, and 100W Epistar should all be the same at each kelvin rating.
Minimalistic multichip DIY LED build
- Thread starter maglofster
- Start date
Just keep in mind due to the slightly different blue bases the Epileds spectra will not be the same as the Epistar based chips. The phosphor is the same so the 20W, 50W, and 100W Epistar should all be the same at each kelvin rating.
Well if they have different blends of phosphors available, then I don't see any reason why a 5500k with a red emphasis wouldn't be ideal. Certainly, we don't want a yellowish tone.
If we can get this, why not:
1) 445nm x 20
2) 5500k reddish x 20
3) 420nm x 20
4) 455 nm x 20
5) 15000/20000k x 20
If we go with a 5500/20000 version, it should come out close to the original 10000/15000 version at full poke.
Here's a question: do they offer a 10000k/14000k phosphor with a red emphasis?
If we can get this, why not:
1) 445nm x 20
2) 5500k reddish x 20
3) 420nm x 20
4) 455 nm x 20
5) 15000/20000k x 20
If we go with a 5500/20000 version, it should come out close to the original 10000/15000 version at full poke.
Here's a question: do they offer a 10000k/14000k phosphor with a red emphasis?
Looks good:
1) 20 x 445nm
2) 20 x 5500k reddish
3) 20 x 420nm
4) 20 x 455 nm
5) 20 x 15000 or 20000k
I slightly favor this option because of less blue.
1) 20 x 5500K with a red emphasis
2) 10 x 445nm, 10 x 455nm
3) 20 x 14000K
4) 20 x 420nm
5) 20 x 10000K
Though, lets wait for the spectral plots to add foundation to our final decision making.
Ron
1) 20 x 445nm
2) 20 x 5500k reddish
3) 20 x 420nm
4) 20 x 455 nm
5) 20 x 15000 or 20000k
I slightly favor this option because of less blue.
1) 20 x 5500K with a red emphasis
2) 10 x 445nm, 10 x 455nm
3) 20 x 14000K
4) 20 x 420nm
5) 20 x 10000K
Though, lets wait for the spectral plots to add foundation to our final decision making.
Ron
Last edited:
The "dream ratio" uses all 45mil Epistar LEDs. I know that much. Some of the low temp data did not indicate chip size. I'll get clarification.
Lassef
Member
It means that the first chip can be run up to 1400 mA/leg what I understand?
Sincerely Lasse
Lassef
Member
I think this is very important if the ratio should be 1.5 white - 1 blue and the some of the whites has low K.
In your configuration - the 420 will probably not "pop-up" because the fluorescence colors will be washed out of the other reflective wavelenghts which are available in quantity. You will have "pop-up" if you only run the blue emitters but as soon as you mix in the low K white the "pop-up" colors will disapear for you eyes. This happens also for the high K white when the intensity rise - I can show that with images
If you going this way I think that you need to change the blue configuration to be around 465 - 480 nm in order to compensate for too much yellow.
If your aim is
I think it is better to work with a true RGB chip and a DMX contoller in this case.
Sincerely Lasse
Last edited:
TropTrea
New member
Wow a 1.5 White to 1 Blue ratio sopunds more like a fresh water tank unless your using a lot of very high Kelvin Whites to the point you don;t even need the blues. I run all my tanks between a 1 watt white to 2.5 watts blue to even 3.5 Watts Blue. But I will say at the 3.5 Watts of blue there is less snap than there is at 3 Watts.
In your configuration - the 420 will probably not "pop-up" because the fluorescence colors will be washed out of the other reflective wavelenghts which are available in quantity. You will have "pop-up" if you only run the blue emitters but as soon as you mix in the low K white the "pop-up" colors will disapear for you eyes. This happens also for the high K white when the intensity rise - I can show that with images
One bad misconseption is that all florescent colors are activated by light in the 420 to 450 nm range. While it is true that many pigments are activated in this rang some are activated at longer wavelenghts up to 600 nm and some by shorter wave lengths down to 400 nm.
Another misconseption is that a little white light will take the snap out the colors in your tank. This is only true with an unbalanced white light source that is in excess of the blue's and greens. On the other end of the spectrum do yuou want your bright red Clown fish to look dark marron or even black?
Actualy the Royal blues with a 450 nm peak wavelenght will work to reduce the Yellow. The biggest problem is with most of the white LED's they balance nicely at different ratios between the Blue and Red but they lack in the greens. The Cree Neutral Whites are the only white LED's that have a good amount of green. Higher K rated LED's are reducing the Red to up the Blue which actualy shifts the yellow orange spectrum, to your eye sensors.
I
Dont forget that specific Red Green and Blue Chips create three distinct spikes at three different wavelenghts. If light is needed at another frequency between these peaks it simply is not there. Unles you want to go with a about 12 specific LED's at different wavelenghts you much better off simply balancing your whites.
Sincerely Lasse[/QUOTE]
I hope all understand that the second chip isn't meant to be driven at 100% on all channels. Also keep in mind the general rules of thumb for LEDs assume lower Kelvin whites than most of the whites in this build. A cool white Cree only goes up to 8000K. We have two legs that are well above this and should provide some additional blue.
We also have the ability to drastically change the color ratio with the drive currents. I can power the white strings at 1000ma and power the blue strings at 2000ma and have a 4:3 ratio of blue to white light. Lower the power on the warm string and you can easily get to a 2:1 blue to white ratio. As I stated above I doubt the need for such a high ratio with the higher Kelvin whites.
Couldn't agree more. We may see it as a nice white but we may not be providing the peaks necessary for photosynthesis. The spectra of a RGB emitter is very different than the plot of a "true" white LED.
Last edited:
BeanAnimal
Premium Member
Lasse,
Respectfully, I am not sure you understand the difference between an LED and a "full spectrum" light source.
An "RGB" source DOES NOT create full spectrum light. They can't be used as "grow lights" or "aquarium lights" because they do not contain the spectral peaks that we need.
An "RGB" LED array produces as single peak of RED and single peak of BLUE and a single peak of GREAN. The intensity of each is controlled to TRICK the cones in our eyes into seeing "full spectrum" (gamut issues aside). The "full spectrum" DOES NOT EXIST! This is very important to understand.
The chips we are talking about here work the same way (tricking our eyes and brains), except that our choices of R G and B are based on emitters that have wavelengths that grow coral. The downside is the color rendering. That is, what we see and what the "coral sees" is not the same thing. The RGB chip is simply the rong wavelengths, no matter what it "looks" like.
So BACK TO THE POINT:
(Again) many of us are looking for a chip that will create a SINGLE POINT source of light that is somewhat controllable, being both able to grow coral AND be pleasing to the eye in different scenarios. Part of that goal is to be able to vary the light over the course of the day, or fine tune it to ones preference. This has been articulated at least 10 times in the last 3 pages.
TropTrea
New member
You can find these plots right on CREE's pages. I realy love there Neutral Whites roughly a hair under 5,000K. It gives a real good balance against there Blues and Royal Blues.
The problem here is it will vary from each manufacturer and there are a lot of different spectra that may render to a 5000k white. The reason the Cree cool white has such a high lumens/watt is because they provide so much energy in the green spectrum. This increases lumens at the cost of color rendering be minimizing the warmer end of the spectrum.
Lassef
Member
You wrote in post 1778
I answered that if this is your aim - when it will be better to do that with a normal RGB chip and a DMX controller. In this sentens I can´t see a word about growth, or anything about other things according to light. I asked if that was yor aim - to be able to replicate dawn, dusk, mid day, reddish skies- and if it was - you can do it with a RGB chip.
In a lot of post here I think that I have shown that I have some understanding on the issue of growth, photosynthesis and other things relating to the biological function of the light. Not in one case in these posts, I have advocated the use RGB but now I asked you if your goal was to recreate dawn, dusk, mid day, reddish skies (as you state in your post) and it is easy to do that with a RGB chip and a DMX control.
I know very well how an RGB chip is constructed and the difference between light sources that have a full spectra and light sources that do not. I am very well aware of what is required of the light at the biological level. I think it is a bit insulting considering what I have written earlier in the thread when you try to claim that I do not understand this.
What I tried to say was that if you want this light effects (thats for your eyes more than for the corals) its easier to create them with a RGB chip than trying to incoperate it together with your main lights.
Sincerely Lasse
Last edited:
Check out the spectrial plots of bridgelux 45mil 6000K and Epistar 35mil warm white 3100K:
https://picasaweb.google.com/103048473220054692919/August82012?authuser=0&feat=directlink
https://picasaweb.google.com/103048473220054692919/August82012?authuser=0&feat=directlink
bhazard451
New member
You could easily run a strip like this for rgb aesthetics.
http://www.aliexpress.com/product-f...150-Led-30-leds-Meter-24-Key-wholesalers.html
This wouldn't look very natural however, and wouldn't benefit coral in any way. It would destroy the benefit of using a multichip as a single point source as well.
Yes, they can offer the higher K temps, like 10000K, with red, but it will add green as well. Let me verify a few things first.
Ron
We would prefer to use only one kind of chip throughout our tank however.
In the example I am leaning towards we have ample light for photosynthesis in the blue range. Both channels of 10K+ white should provide significant spikes in the blue range, in addition to the 2 channels of blue LEDs. Warmer whites will still provide some blue as well. With a reddish tint a warm white will also provide peaks near the secondary excitation point at the red end of the spectrum. In addition this channel can be used independently for a dawn/dusk effect while still giving some usable light to the tank. A seperate RGB emitter would not only compromise the reason for going with 1 multichip it would also only be for aesthetics. Hopefully we will be able to satisfy aesthetics and photosynthesis with this proposal.
I think all of us need to take this stuff for what it is. We all have varying opinions and we are having a frank discussion. I dont think anybody is attacking anybody else. Keep in mind we are talking about lighting for a fish tank.
In the example I am leaning towards we have ample light for photosynthesis in the blue range. Both channels of 10K+ white should provide significant spikes in the blue range, in addition to the 2 channels of blue LEDs. Warmer whites will still provide some blue as well. With a reddish tint a warm white will also provide peaks near the secondary excitation point at the red end of the spectrum. In addition this channel can be used independently for a dawn/dusk effect while still giving some usable light to the tank. A seperate RGB emitter would not only compromise the reason for going with 1 multichip it would also only be for aesthetics. Hopefully we will be able to satisfy aesthetics and photosynthesis with this proposal.
I think all of us need to take this stuff for what it is. We all have varying opinions and we are having a frank discussion. I dont think anybody is attacking anybody else. Keep in mind we are talking about lighting for a fish tank.
bhazard451
New member
That Neutral White Bridgelux (and their WW too) would be something I would use mixed with 1.5/2x royal blue in a 3w led build with a few red, green, and 420nm leds added. It ends up looking really good.
That WW Epistar may be better off for chip #2 though, if we are adding something like 14k as well. Their 14k epistar and 3k epistar legs both running around 130 lm would even out the spectral plot a bit.
I agree on a red push too, not yellow. There will be plenty of yellow to go around.
