Full Spectrum DIY guide

[BeanAnimal]

Not only does the electrical efficiency come into play, but so does the distance from the tank and the use (or not) of optics and the actual wavelengths of LEDs in question.

 

[BeanAnimal]

At this point I really don't care about anything you have to say, you just sit there in your little world thinking your a saint. Far as I'm concerned you are ignorent and since you can't count, or spell or edit your own posts and are so concerned with mine the "ignorent" count is now 2 good day sir

You may want to make note that the proper spelling is "ignorant". I attempted to kindly answer your questions. I am sorry you feel the need to use name calling and personal attacks, but it is not fair to me or the others that are participating in this thread. No good ever comes of it.

 

[BeanAnimal]

L X W / +-16
48x24/16=72

72*3=216

So you would need 216 watts of led lighting for optimal growth. whether or not you use a 3 watt chip or a 10 watt chip,50,100 so on so forth as long as you achieve that 200 range you should be good, id use a optic if i were you.

I would also note (in addition to what was already posted) that that "rule of thumb" was likely based on the 2:1 RB/CW ratio that has been the primary setup that most DIYs and fixtures have used. In context to this thread, that 200W may become 300W or more based on the need for significant amounts of other colors tha are needed for color rendering but not growth. Throwing the multi-chips into the mix makes things even more confusing. Not having run both, I would not know how to compare them, let alone follow a rule of thumb.

I am still firmly staddling the fence between 3w discretes and 100W custom multi-chips.

 

[BeanAnimal]

If you can't understand what I'm asking about here than I don't know what to tell you.
1)Do you feel or think that a single white 10-12k will have more spectral wavelengths than say a single r g or b?

A 10K or 12K LED is built from a blue (or other) LED coated with phosphors. Some of the underlying color shines through as well as the spectral output from the phosphors. The "10K or 12K" rating means that the output somewhat looks like an ideal 10K or 12K full spectrum source. Without looking at a spectral plot, it is hard to see what peaks are in there but due to the phosphors, there is certainly more spectral diversity than a mix of single wavelength R G and B emitters.

2)Are wavelengths specific levels of color present and spectrum is the collection of those wavelengths nessesary to make a secondary color?

A wavelength is a single frequency of light, measured in nanometers. It is part of the electromagnetic spectrum just like a single radio frequency, microave frequency, etc. So yes, multiple wavelengths are present in a "spectrum".

3)Would you agree that whites have other wavelengths required to make white.

Yes, "White" light is the combination of many wavelengths. The subject is a bit more complicated than that, but think about how a prism works. If you don't know (it is beyond easily explaining here) it would be a good starting point.

4)is it true that wavelengths can differ from bulb to bulb depending on source, how it was manufactured and so on?

Yes, if we take a pile of "14K" bulbs from various vendors, they will all be comprised of different amounts of various wavelengths. You need to read about CRI (Color Rendering Index) and CCT (Corrected Color Temperature) to understand the Kelvin ratings assinged to lights. They really have nothing to do with the actual wavelengths and more to do with the way the light looks and/or is compared to the "ideal" black body color temperature.

5)Would you agree that it is not safe to assume that any given white led can provide the nessesary full spectrums to grow corals on thier own, but minus uv is possible?

Becuase a "white" led could be comprised of just about anything, it is safe to say that alone they may not be suitable to grow coral.

This is all I was ever trying to get at. It seems to me that if a white LED's spectral plot shows peaks in blue, red, yellow or orange that those colors are there but may not be visible in thier true colors as they have been mixed to make a new (visually) color, correct or no????

They are in fact visible becuase they are contributing to the final color rendering you see. The ratios of those colors determins not only the "temperature" of the "white" but also the way certain objects look under it. So to go back to the differences between two 14K sources that are side by side. Both may "look" the same color, but object of varying color will display differently under them. This is again explained by understanding CRI and CCT. So a good starting point would be using google to read about CRI and CCT. I am not avoiding explaining it, you would be better served visiting several web resources and reading.

 

[jerpa]

I am still firmly staddling the fence between 3w discretes and 100W custom multi-chips

I stood at a local LFS staring at the ReefTech fixture the other day thinking the same thing. I can't seem to find the definitive specs on the LED's included but it is probably the best discrete emitter design I've seen from an aesthetic perspective. I tried to talk to one of the employees but he had no interest in telling me anything. Do they utilize an MCPCB? It kind of follows your line of thinking with regard to the number of emitters. I can't imagine you'd have to run anywhere near full power with ~70 leds per fixture or whatever it is but using that many emitters gives you lots of color options.

 

[BeanAnimal]

I'll quit with what you call "personal attacks" if you quit trying to be my damn teacher! I came here for led info not grammar lessons. Stick to the topic and we would have never had an issue, you could have from the beginning stated you didn't understand me and not critique my typing and spelling. I'm educated and I can re read my own post and figure it out just fine.

A word of friendly advice. You may want to consider growing a slightly thicker intenet skin my friend. If you continue to post in this manner everytime you do not agree with something, you are going to end up very quickly banned and I am sure that is not what you want.

Calling people ignorant jerks and jackasses and telling them to shut up don't bode well with the moderators or those on the receiving end of the tirade, especially when your anger is a result of somebody kindly asking you to use punctuation to clarify your questions.

 

[BeanAnimal]

I stood at a local LFS staring at the ReefTech fixture the other day thinking the same thing. I can't seem to find the definitive specs on the LED's included but it is probably the best discrete emitter design I've seen from an aesthetic perspective. I tried to talk to one of the employees but he had no interest in telling me anything. Do they utilize an MCPCB? It kind of follows your line of thinking with regard to the number of emitters. I can't imagine you'd have to run anywhere near full power with ~70 leds per fixture or whatever it is but using that many emitters gives you lots of color options.

Jeff is the only guy that really knows anything about them. I agree, they by far have the best color rendering I have seen. In fact, it really is the only fixture I would consider viable for me. A bit (a lot) more than I want to spend.

There are no optics and yes each section is fitted on an MCPCB. Did you notice any disco effect?

 

[jerpa]

Jeff is the only guy that really knows anything about them. I agree, they by far have the best color rendering I have seen. In fact, it really is the only fixture I would consider viable for me. A bit (a lot) more than I want to spend.

There are no optics and yes each section is fitted on an MCPCB. Did you notice any disco effect?

Yeah I believe they're $1200 each correct? I would need two and I just couldn't justify that. With the prices on the new XT-E's it is more than a bit much IMO. I could build a star based system with 10mm stars and get the spacing. The wring would be a nightmare though and it wouldn't be pretty. I could do without individual control of each LED and lightning, clouds, etc. I would want sunrise sunset control and individual channels for each color. The somewhat rocky road the company has followed also concerns me.

Oddly enough it seemed there may have been a bit of disco on the fixture up front. The one over the huge tank seemed like it had none. Perhaps intensity settings and mounting height changed it. I also liked the color of the large tank more than the smaller one. I'm not sure. I felt like my eyes were playing tricks on me after staring for awhile.

 

[BeanAnimal]

I think they are $999 now. I prefer the color of the tank close to the highway (that is the big one I think). It is the older fixture but as you note, likely just setup differently colorwise. I am sure that if you give Jeff a call he will arrange to demo both the old and new fixture for you.

I was about 99% done choosing my 3W layout and started designing my MCPCBs when I became interested in the multi-chip options. At this point I have no idea what I want to do.

 

[dsdaley77]

Thank you, I appreciate you talking the time to answer my question or steer me in the right direction. I was never mad, just annoyed and could care less if I get banned from somewhere online, there are plenty of places to get info. But I will say this based on the answers you gave to my questions, we were in agreement the whole time

 

[dsdaley77]

At this point I have no idea what I want to do.

do have room on that boat? I just want to wake up and find a mysterious fixture, wreaking perfection sitting above my tank

 

[Tickle]

I think using this formula without looking at the efficiency of the specific LEDs in question is kind of silly. Your using a formula made for high end cree and bridgelux emitters. Multichips are always less efficient than 3w designs and often use diodes that are far from the most efficient available. It's not all about the watts so while that might be a good startimg point a lot of it will depend on what specific LEDs you go with.

I think most people here are using the cree, but this formula has been applied to the multichips more than you would imagine, which is also what i based my build off of. My math computes out to 112 i think off hand and i undershot that with 100 watts. I cannot run the multichips at 100% because then i experience bleaching. I also under drive my chips as well. I also stated this was a rule of thumb in a earlier post. if you assemble your fixture with that rule and find you need more than add more but so far in my case it has proven more than adequate.

One thing I would like to point out that Ive seen alot of people get mislead on is that for example a 10 watt multichip is composed of 10 1 watt leds which is incorrect. they are made out of 9 2.5 watt leds. so technically a 10 watt chip could be a 22.5 watt chip, the trade off would be increased heat and decreased lifespan of the chip.

Also that was the whole purpose of this thread which I feel it has kind of drifted away from its original intentions. QUOTE" thought this would make a good area for some of us who make our own, or would like to make our own fixtures. This area could be used to share info on what has worked for you and your tank. It would help others if you were to break down exactly what you used and why, as well as what you changed to achieve a certain color or better growth etc."end quote.

That is what has worked for me and worked well i might add. I will also note that no my build is not full spectral YET. I choose to post my standard B/W setup as a baseline and the rest of spectrum is in the making.I know alot of the big name manufacturers use the 3 watt cree or bridgelux chips but correct me if Im wrong but the but the kessil units are a multichip. http://r e e f builders.com/2011/03/29/dense-matrix-array-led-kessils-inhouse-version-multichip-led/

Im not putting down your statement nor am I brushing it aside.

 

[BeanAnimal]

In the end I will likely end up with a star based prototype using 3W emitters and another using 100W multi-color chips.

I am sure that the 3W setup will give me the color control I want but I fear the disco. I am not sure that hte multi-chip is efficient enough and am very concerned about the choices for color and control but know there will be no disco.

 

[Tickle]

Not only does the electrical efficiency come into play, but so does the distance from the tank and the use (or not) of optics and the actual wavelengths of LEDs in question.

would this not hold true with any fixture built? If I hang a unit 1 inch from the tank with no optics and then hang the same unit 10 inches with 90 degree optics I might get the same par readings with a different look in the tank itself.

Multichips are just as tried and proven as its cree counterpart, I just think cree is a little more known, im not putting down cree in any way shape or form, I was actually on the cree boat before I stumbled into multichips.

The thing that got me about multichips is if I wanted to put 100 watts over my tank I could do it with 10 chips vs 33 chips.

My advantage on the cree chips is that I could then put optics on them to help focus the beam and increase the par at the bottom of my tank without sacrificing spread or getting a disco effect.

advantage on the multichip was purely simplicity and cost. If I choose to use optics i have less optics to buy, i have less solder joints, use less wire and the cost per chip is less than a cree.

PS BEANANIMAL since you cannot receive pm have you read the minimalistic multichip diy thread here on this forum? You do bring up valid points. I feel there has been enough success with this chip to make it a good competitor against the cree and the bridglux and most the questions have been asked and answered in that forum.
I digress though because thats not what this thread is about.

 

[jerpa]

I think they are $999 now. I prefer the color of the tank close to the highway (that is the big one I think). It is the older fixture but as you note, likely just setup differently colorwise. I am sure that if you give Jeff a call he will arrange to demo both the old and new fixture for you.

I was about 99% done choosing my 3W layout and started designing my MCPCBs when I became interested in the multi-chip options. At this point I have no idea what I want to do.

Yeah I understand the feeling. I don't know that there is a feasible way to get as many colors as you may want on a multichip with individual control of each color. If you use multiple multi-channel multichips, say that 5 times fast, with different combinations of emitters you've defeated the purpose of the multichip. I also wonder how something like a 10 channel multichip would blend if you had distinct red and green channels as opposed to a combination of numerous whites and blues only.

If we could get aesthetically pleasing results with just a combination of several whites and blues I think a multichip becomes more feasible. By combining similar blues, such as 410nm and 420nm, 5 channels becomes more reasonable.

Then we get into the efficiency of some of the Epistar and Epiled based emitters versus Cree.

So yeah I mean that is all there is to it.

 

[Tickle]

Yeah I understand the feeling. I don't know that there is a feasible way to get as many colors as you may want on a multichip with individual control of each color. If you use multiple multi-channel multichips, say that 5 times fast, with different combinations of emitters you've defeated the purpose of the multichip. I also wonder how something like a 10 channel multichip would blend if you had distinct red and green channels as opposed to a combination of numerous whites and blues only.

If we could get aesthetically pleasing results with just a combination of several whites and blues I think a multichip becomes more feasible. By combining similar blues, such as 410nm and 420nm, 5 channels becomes more reasonable.

Then we get into the efficiency of some of the Epistar and Epiled based emitters versus Cree.

So yeah I mean that is all there is to it.

I agree with you 100% thats why ive come to the conclusion that I will use crees on a separate string to help get that color blending, in my short experience the multichips have proven to me that in a standard 455/14k 1:1 ratio life will not only be sustained but will flourish. So as a supplement I will run the crees in various spectrums to achieve my full spectrum and get the pop in my tank that I like.

 

[jerpa]

I think most people here are using the cree, but this formula has been applied to the multichips more than you would imagine, which is also what i based my build off of. My math computes out to 112 i think off hand and i undershot that with 100 watts. I cannot run the multichips at 100% because then i experience bleaching. I also under drive my chips as well. I also stated this was a rule of thumb in a earlier post. if you assemble your fixture with that rule and find you need more than add more but so far in my case it has proven more than adequate.

One thing I would like to point out that Ive seen alot of people get mislead on is that for example a 10 watt multichip is composed of 10 1 watt leds which is incorrect. they are made out of 9 2.5 watt leds. so technically a 10 watt chip could be a 22.5 watt chip, the trade off would be increased heat and decreased lifespan of the chip.

Also that was the whole purpose of this thread which I feel it has kind of drifted away from its original intentions. QUOTE" thought this would make a good area for some of us who make our own, or would like to make our own fixtures. This area could be used to share info on what has worked for you and your tank. It would help others if you were to break down exactly what you used and why, as well as what you changed to achieve a certain color or better growth etc."end quote.

That is what has worked for me and worked well i might add. I will also note that no my build is not full spectral YET. I choose to post my standard B/W setup as a baseline and the rest of spectrum is in the making.I know alot of the big name manufacturers use the 3 watt cree or bridgelux chips but correct me if Im wrong but the but the kessil units are a multichip. http://r e e f builders.com/2011/03/29/dense-matrix-array-led-kessils-inhouse-version-multichip-led/

Im not putting down your statement nor am I brushing it aside.

I wasn't saying it wouldn't work I just have a hard time with rules of thumbs for LED's when there is such a discrepancy between emitters. Some 10w chips can be run at 25w others cannot. Some 100w can be run up to 250 watts, others cannot. Some are based on epileds, some on epistar, and others on undisclosed brands.

I hope I didn't give the impression I'm against a multichip. It is actually my ideal solution assuming I can get a color I like and accurate color rendering while still maintaining a good degree of efficiency. Yes the kessils are a multichip design comprising at least 5 different diodes. They are 2.5 watt LED's produced by their parent corporation DiCon.

I would contend that what we are now talking about in the last few posts is absolutely in line with the nature of the thread as indicated by your post.

 

[Nanofreak79]

Jeff is the only guy that really knows anything about them. I agree, they by far have the best color rendering I have seen. In fact, it really is the only fixture I would consider viable for me. A bit (a lot) more than I want to spend.

There are no optics and yes each section is fitted on an MCPCB. Did you notice any disco effect?

Excuse me if I'm incorrect here, but aren't the fixtures using xp-g and xt-e Cree? Also I thought they were using optics? Maybe I'm wrong. They seem very expensive. I'd like to see someone talk about doing a DIY similar to that fixture, with say an arduino base system to control it. That would be more interesting.

 

[PolyReef]

1)This is funny.....
2)Best advertisement I ever saw in my life.....
3)I think Google is a better option....
Happy racing ....sorry reefing

 

[jerpa]

Excuse me if I'm incorrect here, but aren't the fixtures using xp-g and xt-e Cree? Also I thought they were using optics? Maybe I'm wrong. They seem very expensive. I'd like to see someone talk about doing a DIY similar to that fixture, with say an arduino base system to control it. That would be more interesting.

I'm debating between going this way or multichip, much like Bean. I've had trouble finding concrete specs on the lights since they don't have a company website and some sites seem to list the old specs for the new light. I didn't see any optics on the fixture I saw.

1)This is funny.....
2)Best advertisement I ever saw in my life.....
3)I think Google is a better option....
Happy racing sorry reefing

If you have a desire to participate in the thread please do otherwise your comments are irrelevant. I've only seen one link to a commercial fixture in this thread. Now who posted that. Hmmmm. Oh wait! It was you!