an open ended discussion in Full Spectrum LED concepts

Discussion in 'DIY' started by HiFidelity, Jun 29, 2013.

  1. HiFidelity

    HiFidelity Guest

    Hi Gang,

    I will begin with addressing the title and overall objective

    Title:
    -Open ended discussion for any and all to chime in with any ideas and/or first hand experiences/experiments.
    -Full Spectrum LED lighting; for all intended purposes we will be strictly focusing our attention on the light spectrum absorbed/used up by corals under sea water, not to confuse the conventional meaning of "full spectrum" and that is the full range of sunlight. It seems that achieving Full Spectrum output has quickly become the holy grail of LED lighting, you can notice a trend with commercial brands over the past couple of years shifting from a blend of Royal Blue & Cool White to Full Spectrum, employing the use of yellows, reds, greens as well as violets. This combination was rarely seen in older production, and that's exactly the type of light configuration we are going to discuss.

    Objective:
    -Explore and expand on the most up to date concepts detailed or described by others from this or any other forum on the web.
    -Explore newest available (for purchase) products on the market in comparison to the pool of successfully employed current technologies.
    -Address the advantages and disadvantages of full spectrum vs. other common configurations.
    -Create a platform on which we can expand and elaborate collectively for the mutual benefit of the club without an end date or timeline so the topic can hopefully evolve as we go along.

    Intro:
    I have found our little haven of awesomeness we call BAR to be one of the most well versed databases of knowledge (discussed on the forum and through the events & meets) when I started with my current tank I found BAR to be the one-stop-shop to answer some of the most important questions in the hobby, and with great accuracy. Therefore I would like to contribute to this big ball of greatness with some info for all members to enjoy, specifically on a topic that is very important in our world of Reef Tanks (of course with some help).


    So I am one of those people who's brain works like a jig saw puzzle, I can't really approach things simply out of curiosity or sheer desire, I must first assemble an ample amount of knowledge pre-hand before I delve into anything new and needless to say that can be infinitely complex when we're talking LED's. So basically I'm constantly researching and studying available data from previous studies and experiments from any and every source I can find on the net as well as others in the hobby.

    Without too many confusing details I'll begin with a basic full spectrum configuration;

    2:1 RB:NW (or close to it)

    Neutral White (NW) comprising of a spectral range of 4500K-5000K

    Supplement Warm White (CW) lighting of a spectral range of 3000K-4000K

    Supplement Violet, spectral color of 400-420nm

    Supplement lighting covering 400nm-700nm to target the range utilized through Photosynthetic Absorption (OCW LEDs come to mind)

    Brand & model of LED based on tank dimensions, depth, type of inhabitants as well as personal preference

    These figures and ratios are in my opinion targeted to provide good light output for photosynthetic absorption of the zooxanthellae, specifically in relation to the chlorophyll a and chlorophyll c pigments, simply put "to grow healthy corals"
    There are also carotenoids which are the second reason for the 400nm-700nm supplementation, this chart paints a good picture;

    [​IMG]

    Of course we need a useful graph that shows the corresponding colors for each of the numbers mentioned;

    [​IMG]

    I specifically did not list LED models or number, but rather I wanted to quickly include a baseline ratio of LED colors for whoever stumbles across this thread through search or otherwise. This way they can get the first question (what color LEDs should I have?) about full spectrum answered in the first post. These recommendations are subject to change as the discussion evolves since this should be a collective effort so hopefully we form a more accurate hypothesis.

    I have soooo much to share with you guys; data, graphs, studies, and of course other discussions but I just realized that this is becoming very lengthy already so for the sake of avoiding info overload I feel this is a good start.

    So enough from me, how about we get your feet wet with a couple of hot questions being widely posed and debated recently;

    Full Spectrum vs Dominantly Blue Lighting (on the left of the spectrum, ie; the 20000k look), which is really better for optimal coral growth? and why?

    Remember that a full spectrum system would employ a lot of neutral white LED's (like the ratios listed above) and blue-look would be the often commercially favored 2:1 RB:CW blend.

    LED Sources:

    http://www.ledgroupbuy.com/
    http://www.rapidled.com/
     
    Last edited: Nov 17, 2013
  2. BAYMAC

    BAYMAC Guest

    Do corals produce their own carotenoids via photosynthesis? I thought they didn't? I know just who to ask though....
     
  3. HiFidelity

    HiFidelity Guest

    Quick answer is no, Corals do not produce their own carotenoids, they are either directly accumulated from food or partly modified through metabolic reactions.

    Photosynthesis is a job performed purely by zooxanthellae to produce carbohydrates, the coral's energy primary energy source.

    To further elaborate on Photosynthesis (since it's very important for this topic) Coral's don't do the photosynthesis, their symbiotic zooxanthellae do the job.

    Zoox take up CO2 which comes from bicarbonate ion diffusion (through cellular layers) then photosynthesis happens within zoox which creates carbohydrates, which are then used to create ATP (adenosine triphosphate) which is then used to transport bicarbonate ions, which eventually leads to the deposition of calcium carbonate (aragonite) and then you have skeletal formation.

    This process stops within 10 minutes of your light shutting off, this happens because about 95% of the energy is generated by photosynthesis in the zooxanthellae. The process has been scientifically called "light-enhanced calcification"

    haha sorry for the lengthy answer, I felt that I had to elaborate on how & why photosynthesis occurs in corals since the core of this topic is photosynthesis and how we can enhance it. Thanks, your question pointed me to a mistake I overlooked in my post where I accidentally called carotenoids "pigments" that's corrected now. :)
     
  4. rygh

    rygh Supporting Member

    Cool. Always a fun topic.

    One issue:
    > Neutral White (NW) comprising of a spectral range of 4500K-5000K
    > Supplement Cool White (CW) lighting of a spectral range of 6500K-10000K
    They do not really complement/supplement each other.
    Compare the spectra of each in detail.
    Essentially the cool white is a spiky subset of the Neutral White. Less on the upper end, and a far larger royal blue spike.

    Another issue:
    The red upper end is tricky. Good for PAR, looks nice, since reds on a lot of tanks are weak. Somewhat good for corals, if limited.
    But unfortunately also really good for algae.

    Personal opinion of the day. (It changes)
    You want to start with Neutral white, for high CRI and wide spectral range. No cool whites.
    Add lots of royal blue, as normal.
    Add violet (near UV) instead of red. Looks similar to the eyes, but better coral PAR, and less algae growth.
    So 1+2+1, in a tight cluster, perhaps on the same MCM.
    But have all 3 on separate channels, for tweaking.

    I am still uncertain on adding cyan (blue-green) to fill in the missing spectrum. At this point I think keeping it simple
    is more important than adding it, since I have a lot of good looking greens and blues, with no cyan supplement.

    I tend to like 420-430nm violet better than the 400nm UV range.
    It helps to replace the red. Less of an issue with acrylic blocking it.
     
  5. JAR

    JAR Supporting Member

    I am seriously considering LED's for my reef.
    I need to know if I can grow all the stuff I am currently growing with my 2x250 MH set up.
    Are we there yet?
    Will LED grow SPS coral like a MH?
    I am planning to go to the Kessil Tour to learn as much as I can from them.
     
  6. Coral reefer

    Coral reefer President

    Yes Jon, you can. We are there. Several options. AI, radion, kessil all will work well. I can help you build a custom one if you want.
     
  7. denzil

    denzil Webmaster

    Works for me. :)

    However, I have yet to do a controlled experiment. Would be awesome to conduct one.
     
  8. bayview

    bayview Guest

    2 watts per gallon seems about right; Lenses can burn, spectrum can burn, lots of variables ...

    I have tried and made some various lights. Some are good at growing coral some are good at viewing coral

    50/50 neutral white royal blue IMHO Fugley grows lots of algae ...
    [​IMG]

    Full spectrum custom;
    [​IMG]
    [​IMG]
    [​IMG]
     
  9. bayview

    bayview Guest

    cool white + Blue Led 60 degree lenses great for growing coral just don't toast them much better color less Windex look than above
    [​IMG]
    Full spectrum 405nm 450nm 470nm 3500k 660nm 120 degree for viewing, to much uv / red o have on all the time ... or have to dim
    [​IMG]
    Both 150 watts total over the 120
    [​IMG]

    [​IMG]

    [​IMG]
     
  10. HiFidelity

    HiFidelity Guest

    Hi guys, I'm very excited to come back and read everything in detail...

    Before the forum went down I had started working on a chart in photoshop where I was layering different LED spectrums onto it so we can have a visual of all the colors available, their range and where they fall on the spectrum.

    in the meanwhile HAPPY INDEPENDENCE DAY EVERYONE :D
     
  11. decadrum

    decadrum Guest

    My cruuent tank has been running CREE LEDs from its inception (approx 2 years). The tank is 34" deep and after some trial and error, I have been getting great growth with everything I put in. There are 96 LEDs with 40 cool white, 30 royal blue, 13 red and 13 green. I chose the make up based on several different articles about what wavelengths corals respond best to. Unfortunately, I have no idea where to find them now.

    [​IMG]

    [​IMG]
     
    bayview likes this.
  12. sfsuphysics

    sfsuphysics Supporting Member

    IMO, when I see the term "Full Spectrum" with regards to LEDs I get a twitch I can't control. Full spectrum seems to be one of those buzz words that a lot of LED makers have started touting but what people who are into LEDs use it as makes me think they're not really sure what it is either.

    Full spectrum: Presumably meaning like actual sunlight... fairly well represented as an equal amount of power at all wavelengths, except for the far violet which it tends to drop off, but still it's somewhat strong (maybe 30% less energy), and to a lesser extent reds (drops off maybe 10-15% compared to say yellow), this is why purple looks really dark to us and also why reds have a "dark" feel to it... our eyes simply evolved to not be super sensitive to those colors (more so with the violet part than red).

    What we've been using in the hobby: Not anywhere close to full spectrum. Often it will be a very bright blue spike with a lot of "background" at other wavelength, the amount of "background" will be dependent upon the color temperature... usually.
    From Sanjay's reports
    20kK radium - big spike in the blue.. but it's a 20kK bulb, so it's expected
    [​IMG]

    10kK XM - Big spike in blue, and a lot more "background" but overall still not terribly much compared to the blue
    [​IMG]

    6500K Iwasaki - Now this is the only one that comes anywhere close to a "full spectrum" IMO... however anyone who's used these (ok except for one guy in Texas) usually don't like them as far as color
    [​IMG]

    So if we aren't really after full spectrum, what are we doing?
    Well it seems that most people throw in a red LED and perhaps a green or cyan colored LED and declare that to be full spectrum. Arguments I've heard is it brings out more "true" colors, or they have a photosynthetic spike near there, don't like the "Windex" look. All perfectly fine argument lets take this one at a time
    Reds pop/True color: Well you put something under more yellow light than yes it will look more yellow, bathe it in blue light and assuming it has no fluorescence then it will look more blue, if you're going to pay attention to the red and the blue and neglect all those other tiny wavelengths you're not going to get a true coloration anyways. This is an argument of aesthetics, which IMO is perfectly valid... at least for you. What you think is pretty is perfectly fine, what I think is pretty might differ.
    Photosynthesis in reds: This is absolutely true, that doesn't mean that they absolutely NEED to use the photosynthesis that's at red. Corals have grown for decades under light that has a lack of red light, they've been just as colorful and just as crisp as before. Here's the kicker, those graphs of chlorophyll you see, isn't a representation of how strong the light needs to be at those wavelengths, it just shows where the most efficient conversion/absorption of energy is. Here's a popular graph
    [​IMG]
    While yeah there are spikes at the red part, there's also in the blue, so you'll still get absorption with blue bulbs, in fact this is why our corals grow fine without that red part of the spectrum that everyone clamors after. So your corals will get by with just a little bit of red just fine as it'll be very useful at absorbing it.
    Windex look: Again this is an argument of aesthetics which is perfectly valid. IMO, the "Windex" look comes when you stick with a single color type, i.e. only blue, I've seen it with blue T5 bulbs as well, it's not only something that LEDs create. But take actinics, dark purple almost can't see much light, except through fluorescence again our eyes are not very strong at those wavelengths so we don't see that light as very bright, but the corals fluorescing does pop out.

    In my experience with LEDs "cool white" will lead to that "washed out" look simply because they have a lot of blue as opposed to other colors(if you've ever bought "cool white" fluorescent bulbs you'll now that sterile mausoleum look), however if you change up those cool whites for neutral whites you have a completely different game. Especially with the latest Cree XTE neutral white bulbs they have an absolute ton of red in them, and those will do more for color rendition too because they give you red along a wider spectrum than say a single red LED which tends to give you a very sharp red spike (660nm is a popular one). And IMO, adding a spike of red does not magically change the light to "full spectrum" in fact it would do the opposite effect because if you're going to argue the blue spikes are not natural, then adding a red spike would be equally not natural.

    Overall though corals are fairly tolerant about what light they receive, although I've seen studies that show corals under only red light will die, however corals under only blue light will thrive, so perhaps there's not much to that "need" of red light. But the best thing to do is to find what you like as far as color spectrum, and if you're unsure, make sure whatever setup you get allows you to dim various channels so you can fine tune your color spectrum, while a lot of fixtures allow you to tune a "white" and a "blue" channel some newer ones are adding violets and other colors to the spectrum.

    My experience is with cool whites and royal blue leds, I prefer a 3:1 (blue to white) ratio or maybe even 4:1 if you're running them at different currents. Its fine when you have stuff that is rally fluorescent however overall it can look washed out. My latest fixture (I think there's a build thread here somewhere) I do 4 separate channels, a 1:1 ratio of neutral white to royal blue, royal blue, 2:1 ratio of blue to royal blue, and violet. The last too channels are run at different power levels so aren't as bright (the violet naturally is like that but it's still only run at 700mA), and the "white" channel is run at around 85% and it does go through various fading as they turn on/off. I find the color is fantastic! However this is an aesthetics argument, so your view might change ;) Corals all seem to grow fine, some not as well as others sure, but you'll get the same thing with MH lamps where you might get a frag from someone that just doesn't do well in your tank, and often curse the water quality or some RTN disease or what not when in fact it might just be the light it's not digging :)
     
  13. HiFidelity

    HiFidelity Guest

    I thought I'd come back with some updates...

    First I realized that I have to read most of the replies again since i'ts been a while since I was on here DOH!

    for now I'll just throw in what I was working on lately (in theory) though I will be experimenting with some other temporary configurations since now I got my hands on 3 Makers LED heatsinks, a bunch ofdrivers and hopefully soon spare time...

    Moving along, so once I have all the necessary parts I thought I'd give this plan a shot;

    [​IMG]

    From left to right;

    1. UV 405-415 1x 3w
    2. UV 420-425 1x 3w
    3. RB 440nm 4x 3w (maybe 5x)
    4. RB 450-455 4x 3w (maybe 5x)
    Green line is NW 4x 3w (maybe 5x)
    White slopes highlight the output spectrum of OCW 3up chips 2x

    So a total of 18 (or 21) per cluster over 18"x18" of tank surface, all 3w & 120 degree optics.
    The objective behind this theory is to cover as much of the spectrum used by corals as possible by varying the UV & Royal Blue spectrums as well as the use of OCW chips (470, 495, 660), the ratios are not static as each channel will be dimmed separately, for example 2 OCW chips over that area will throw a bit too much red & green but once dimmed down the ratio is decreased but by using two chips there will be more spread and even coverage than a single chip at brighter output. Additionally the RB 440, 450-455 & NW will be each soldered onto 3up chips so each chip has one of each of these LEDs.

    That's all for now :)
     
  14. rygh

    rygh Supporting Member

    Not sure all that UV is a good idea.
    Unclear if it is beneficial or damaging. And not the best if you accidentally look at it while working
    in the tank.
    Perhaps consider more of the "True Violet" SemiLEDs ones that peak at 219nm.
     
  15. sfsuphysics

    sfsuphysics Supporting Member

    Hmm the OCW comes with a red LED on the board?

    What type of NW chip are you going with? Because not all NW are anywhere close. Look at the Cree XTE NW (red line in picture) it has a rather large orange/red bump that pretty much makes any sort of dedicated red LED chip obsolete. Yeah not all the power is at 660nm, but easily half of the power is at that wavelength. The benefit is that it frees up "chip space" elsewhere because it does have that red look that people often want and since most people pile on the RB anyways the fact there's not much power in the blue part of the spectrum is irrelevant.
    CreeXTE.png
     
  16. rygh

    rygh Supporting Member

    Partially agree. I think the new neutral whites PLUS the new violets make the reds obsolete.
    The neutral white is good, but the violet fools your eye into seeing more red as well,
    which finishes off the issue.
    If I was building mine again, my leaning would be:
    Get a tri-star, that has 3 leds per star, RB+RB+NW:5K.
    (Steve's LED has nice ones using Phillips Rebel ES, my current favorite over Cree)
    Then add a true violet right next to it, and probably be done.
     
  17. HiFidelity

    HiFidelity Guest

    Mike-
    I completely agree with your comments, specifically with the intended meaning of "Full Spectrum" in this topic. I went ahead and modified the first post to include the definition for "Full Spectrum" within the content of this thread. We certainly only focus on a specific spectrum in the way we use LED lighting in my opinion not to necessarily replicate actual sunlight but to provide the coral with as much usable light as possible and of course we can never deny that we want aesthetically pleasing colors too. In fact aesthetics often are equally important to the hobbyist as quality of light. I'll admit I'm guilty, I'm very particular about how the lights look and do prefer certain configurations over others.

    In regard to colors, I wouldn't say that corals NEED certain colors but if a little bit of red or perhaps green can possibly yield a small gain either in photosynthesis or aesthetics it definitely is worth throwing in one or two more LED's once we've clustered 12, 18, or even 80 LEDs together it's not a big deal to add a few independently controller colors, of course as long as they produce a desirable effect.
    Essentially yes the right blend of Royal Blue with Neutral White will produce the usable "Full Spectrum" for our purposes, but again I would have to say that a very small ratio of colored LED's specially when clustered together adds a small emphasis in the spectrum where Royal Blues and Natural Whites may not be so strong, for example your average Royal Blue LED is in the 450nm range, this is where violets come into play, now this is where there seems to be a lot of misinformation. There is a very wide range of Violet light to be had, adding the right colors can greatly supplement the far left of the Spectrum where corals definitely use the light for growth. Again nothing crazy in numbers like Royal Blues & Whites but a small supplement will stretch the output spectrum a bit.

    Naturally all channels have to be separately dimmed otherwise there will almost surely be too much or too little of something or another resulting in aesthetically undesirable results.

    As far as the chips I'm going with, I'm going to use 4000-4500K Neutral White Bridgelux, unfortunately I can not find an actual spectrogram of this specific chip so I put an estimated curve on the image. I can confidently say that it does not have nearly the spike of the reds in the XTE chip which has me sort of confused, isn't 2600-3700K warm white? I could be wrong but I thought NW is more in the 4000-5000k range!


    Mark-
    You have a good point with blending NW & CW, you are right it isn't necessary because where NW lack in output the Royal Blues provide more than enough thus rendering the supplementation of CW unecessary.

    You've pointed out a few times that Reds can be replaced with UV, while they might look the same to the eye they certainly are not the same kind of light, the photosynthetic action is on opposite ends. Now if this is done merely for aesthetics then yes the red is not necessary anymore when mixing Violets & Neutral White (a little less orange red & more purplish red but close) though if we do this thinking that the coral will absorb the same amount of red then this can not be true, though a very insignificant gain might be had from the Red over Violet theoretically they are not the same light. Essentially I will not argue with you guys about the red, I would probably lean towards removing the Red LED from the OCW chips and replacing it with Violet or better yet Warm White in the 3000k range, so then there is a blend of Cyan, Green & Violet (or low lerd from WW LED) which I am willing to bet is a much more useful output and far less promoter of algae growth.

    In regards to the harms or dangers of UV LED's most of the claims that it will damage your eyes or fry your corals are inaccurate. The wavelengths I selected are out of the damaging (or dangerous) UV range, the ones to worry about are UV-B, UV-C & so on (there are several) and they are all well bellow 400nm (315nm & lower) and for the most part are not something neither us on land nor corals in sea are exposed to in nature, anything >400 is the same as UV exposure under natural sunlight and the ones I selected (405-425) fall in the True Violet range in fact 425 is bluish purple. I was even thinking of adding 1 430nm and putting all 3 on a single chip, beyond that I may even go with 6 UV LED's per cluster and dimming them more so there is more even coverage.

    So now I'm thinking the configuration should go this way.

    1. UV 405-415 1x 3w (maybe 2x)
    2. UV 420-425 1x 3w (maybe 2x)
    3. UV 430 1x 3w (maybe 2x)
    4. RB 440 4x 3w
    5. RB 450-455 4x 3w
    6. NW 4000-4500k 4x 3w
    7. OCW 2x

    The OCW can be modified to eliminate the Red LED and replace it with something more spread out like a warm white LED in the 3500k range (more red but a lot) or Violet.

    The totals are 21 LED's at 63W, this is per cluster and I plan to run 3 of these and dim them way down to increase efficiency, coverage and decrease heat.

    I'm opting to use all Bridgelux and/or Epileds due to the LED design and having the ability to solder on & off chips however which way I want, and the 3up chips for this type of LED are easy to find and very inexpensive.
    The colors are available from many different brand so these ratios will still apply to Cree or Philips but the scale would be lower due to the higher output, but in my opinion more LEDs w/ less power is always better than fewer LEDs running hotter (hence my overkill three 21-LED clusters over 45 gallon tank).
     
    Last edited: Sep 17, 2013
  18. rygh

    rygh Supporting Member

    A clarification regarding the reds:
    My intent was deliberately to fool the eye, and specifically not have the red photosynthetic action.

    The reason is algae growth.
    Near-surface type algae, like hair algae, like red.
    Coral can use the blues more. Sure, needs some red, but the whites should provide that.
    So to emphasize coral growth over algae growth, you skimp on the reds, and crank up the blues.

    I have even semi-scientifically tried this myself as far as the algae part.
    I played around with LED types when doing early algae scrubbers, and the red works WAY better.

    Key : Consider that when looking at the different white LEDs.
    That is why I like the Rebel ES 5K. A lot less energy up in the 660nm range.
     
  19. HiFidelity

    HiFidelity Guest

    [​IMG]

    Just what I have my hands on right now, there are about 50 more LEDs I'm still waiting to get all different colors including what I listed in previous posts as well as some other colors I plan to experiment with o_O
     
  20. rygh

    rygh Supporting Member

    Pretty good sized pile there!!

    Advice for that: Create an assembly line like setup.
    (Although it appears you might be already, since it looks like stars are pre-tinned)

    Other tips:
    Get a permanent black marker, and mark the "-" on each LED.
    Solder red wires to "+" stars, then solder black wires to "-".
    Then after you glue them, you just connect the wires together : Always red to black.
    That is more soldering, but far less chance of mistake, and easy to spot.
    And just as importantly, easier to fix a mistake. Unsoldering two wires is WAY easier
    that desoldering a star glued to a big heat sink.
     

Share This Page