Finally going LED

  • Get the NEW AquariaCentral iOS app --> http://itunes.apple.com/app/id1227181058 // Android version will be out soon!

Ace25

www.centralcoastreefclub. com
Oct 3, 2005
5,753
0
36
www.centralcoastreefclub.com
I have been watching LEDs progress for many years now and have been waiting and waiting until I felt it was mature enough technology to use on a SPS reef tank... well.. this was the year I felt that happened. So I have been researching all the different fixtures out there and DIY methods and finally decided on which route I want to go.

So.. first I ordered a new controller. I went with the Reef Angel controller because it is open source and you can write code for it to make it do pretty much anything you want, especially in the LED area. With the PWM signal and WiFi attachment I can finally do what I have wanted to do for over 5 years now, which is to write software to be able to lookup weather anywhere in the world and have my LED lights mimic the weather. If I set it to read weather from the Great Barrier Reef, I can have it offset the time to match my local time (ie, raining at 2pm there will create rain effects at 2pm on my tank). I plan on having several different "cloud" settings from partial clouds to complete overcast as well as various degrees of lightning storms for the advanced effects. Of course it will do gradual ramp up and down of the blue/white LEDs for a smooth dusk/dawn setting. It will also control moonlights and alter intensity based on the phase of the moon. So that is the controller part of the equation. Total cost was $368 for that part.

Next is the LEDs.. I am going to order from rapidled.com but I am still trying to fine tune my order before I place it. Since this is my first attempt at a MH replacement LED fixture I am not 100% sure on how many LEDs and what colors to go with. They will all be Cree XP-G for whites and XR-E for Blue/Reds. I am not sure about which lenses to go with though nor the exact amount of each type of LED. I am thinking a total of 48 LEDs and 4 meanwell drivers. The DIY kit comes out to around $500 for a 48 LED kit. I will order the heatsink from heatsinkusa.com because they have more sizes and ones that will fit perfectly where I want to place it. I am also going to use my 4 T5HOs along with the LEDs, ATI Blue Plus bulbs, since I am planning on a 1:1 ratio between white and blue/red. One concern I have read about LEDs is the complete lack of UV light from them. They are addressing that issue now with UV LEDs but I think my 4 T5s will provide enough UV to keep the corals happy and healthy. The negative with 0 UV that I have read is UV is used by the coral to help create the "slime coating" they make for protection. To me that related directly to long term health of corals and I think current LED lights without any UV may eventually lead to sickly corals. How long that may take or if it would even happen I have no idea, just a guess from what I have read. Total cost for the LED portion is going to be about $600 making total cost for a top of the line DIY LED fixture w/ controller to be around $1000. Not the cheapest but I should make my $ back within 12 months just on electricity and bulb replacement.

So anyway... this is the start of the project. I should get the controller tomorrow and start programming code for it. Rapidled.com is on vacation until the 29th so I have until then to get my LED configuration worked out before I place the order. I will update this thread as I get parts and start building the light. I am guessing it is going to take about a month before the conversion is completed but I am fine with that. My current lighting is ok now so I am in no rush to put it all together asap. I would rather take my time and do it right.

That's all for now.. :)
 

Amphiprion

Contain the Excitement...
Feb 14, 2007
5,776
0
0
Mobile, Alabama
Real Name
Andrew
Sounds good, so keep us posted.

One thing I have been confused about for some time now is the application of UV LEDs. I don't understand how that could have an effect on corals. Most fixtures, especially ones with some sort of shield, have practically zero UV transmission to begin with. Beyond that, biologically speaking, the coral has to deal with UV wavelengths at a metabolic cost. Almost all data available suggest that corals pretty much always do better without it, too. I'm thinking that it may be the relatively narrow spectrum (especially by comparison to even a 20000K halide) of LEDs, though that is a stab in the dark at best. The UV just seems like a waste to me, though.
 

Ace25

www.centralcoastreefclub. com
Oct 3, 2005
5,753
0
36
www.centralcoastreefclub.com
So.. hours and hours of reading just this week on LEDs.

I found some better red leds here. They are 660nm, ideal for the upper photosynthetic range and something that is severely lacking in the standard XP-G cool white / Royal Blue combos you see on the vast majority of DIY LEDs. The Cree red LEDs are around 625nm, lower than I would like on the red side. That site also sells UV lights but I think I will be running 420nm T5HOs because there is not a 420nm high powered LED. There is also green Cree LEDs but it is very green and not a spectrum really needed for corals. If I think I need green later on I can always add them.

So while my LED setup is still going to be very rigid looking under a spectrometer, lots of strong peaks and gaps, I hope to get all the essentials covered and hopefully a good fill of other colors.

Here is a chart showing the 3 different Cree white LEDs + a Phoenix 14k from data taken from Sanjay's site. I highly doubt the Cree chart is accurate though but it is the only chart out there, but created by Cree. I have a hard time believing it has such a nice smooth curve, I think they averaged the results to smooth out the graph. Anyway, CW=Cool White, NW=Natural White, WW=Warm White. Looking at the chart it seems there should be more than enough color in the middle of the spectrum, much more than a Phoenix or Radium bulb has. The regular blues and royal blues will fill up the missing 450-500nm gap and the red leds will fill up the missing upper end red and the 420nm actinic T5HOs will fill in the missing lower end spectrum.

 

Amphiprion

Contain the Excitement...
Feb 14, 2007
5,776
0
0
Mobile, Alabama
Real Name
Andrew
If accurate, the warm whites mimic noon sunlight fairly well with a decent spectral distribution. I guess I kind of expected that out of a more "daylight" LED, however. Most of the other setups with the usual combos, as you already mentioned, seem to be pretty lacking, though.
 

Ace25

www.centralcoastreefclub. com
Oct 3, 2005
5,753
0
36
www.centralcoastreefclub.com
In regards to UV, this could all be snake oil salesman tactics, but there was a guy at Reef-a-Palooza pushing UV LEDs and even had a UV meter and showed how much UV came off various MH and T5 setups and his UV LEDs he was pushing. While glass shields and SE MH bulbs have UV protection, according to his meter it is not 100%, more like 90%, so 10% still gets through. Which nm of UV that is I have no idea, and what effects UV or lack of UV really has on a coral is still a mystery to me but he did sound convincing. Some google searches on UV for corals showed me the old line of thinking was "UV is required to get color, UV helps the coral make the slime coating and that provides coloring"... but that myth was scientifically proven wrong, there is no pigmentation at all in the protective mucus that corals secrete. So now the thinking is it helps the corals overall health/production of the mucus and without UV the mucus production goes down and makes the coral more susceptible to things like infections. Again, only repeating what I have read and heard, I have no idea how much if any truth is in that.

Oh, and my current plan for LEDs are as follows:
12 - XP-G cool white (1 driver just for these so I can run more power to this string if needed, 1000mA, all the rest max out at 750mA)
6 - XR-E natural white
6 - XR-E warm white
12 - XR-E Royal Blue (460nm)
6 - XR-E standard blue (480nm)
6 - Reds (660nm)
 

Amphiprion

Contain the Excitement...
Feb 14, 2007
5,776
0
0
Mobile, Alabama
Real Name
Andrew
Well, take what you will as far as UV goes. It also depends on the materials the shields are made from. In any case, it won't be something I'll likely supplement any time in the foreseeable future, however. I'm sure my SE lamp is emitting more than what I'd ever want.

That array looks fairly complete, of course with a slight edge toward blue. I know that's what you prefer, though :p:
 

Ace25

www.centralcoastreefclub. com
Oct 3, 2005
5,753
0
36
www.centralcoastreefclub.com
Ya.. your right.. thinking about it more that is a blue setup.. so I will change it a little.

12 - XP-G cool white (1 driver just for these so I can run more power to this string if needed, 1000mA, all the rest max out at 750mA)
6 - XR-E natural white (driver 2)
6 - XR-E warm white (driver 2)
6 - XR-E Royal Blue (460nm) (driver 3)
6 - XR-E standard blue (480nm) (driver 3)
6 - XR-E Green (500nm) (driver 4)
6 - Reds (660nm) (driver 4)

+ 4x 22" T5HO 420nm actinics around the perimeter of the heatsink. That should just above fill in the entire spectrum adding the greens. My goal is to build a light with as much spectrum as I can because with dimming capabilities I can adjust different aspects to then balance out the coloring and PAR to my liking. As far as optics go, actually I think I am going to try it without any at first. I have been looking at pictures with optics and at 14" above the water (what mine will be at) even 60degree lenses were causing bad spotlight effects. They have built in 90degree optics in the LED already with should give a good spread and blend of colors before it hits the water and not have multi-color spots in the tank. With 48 LEDs I should be able to still achieve a good PAR reading, better than my radium, even without optics. I will definately take MH vs LED pictures and PAR readings for comparison. One good thing about having a PAR meter, I can get the PAR of my current setup and when I place the LEDs in I can adjust them to match the PAR the MH was putting out so there is no shock to corals. I am hoping I can get a equal PAR reading while only running the LEDs at 40-50% which would give me room to raise lighting levels slowly.

For the storm/lightning effect I am going to program the controller to randomly swing the power from 20-40% on just the blues for a cloud effect and have the XP-Gs randomly flash on to 75% for 1/2 a second to create a lightning effect. For just a cloudy day scenerio I would run the blues/reds/greens at normal daytime settings but have the 2 sets of different whites randomly dim from their standard daylight setting to 40% power to give the effect of a cloud moving over the tank.

Having a controller that will allow me to do all that is going to be awesome. I have always wanted total control of my lighting, intensity, and spectrum and this is the closest I have gotten yet. I am sure there will be many more advances in the coming years as well.
 

fsn77

AC Moderators
Staff member
Feb 22, 2006
3,076
2
38
SC
Wow, Ace... This is very exciting! I really have nothing useful to contribute, but can't wait to see comparison pix and hopefully video of the storm / lightning effect.
 

leocom2000

AC Members
Dec 27, 2007
434
0
0
I also would like to know how it turns out. I have been looking at something similar to your setup, but it was too expensive for me.
It should be strait forward except for the programming part, but share your problems with us and we will try to help.
 

DoctaQ

all your wheel are belong to us
Dec 12, 2008
1,114
0
0
Boston
heya ace, i know this is your build but if i can make a few suggestions, glad to see you turning to the led side and in style

royal blue should be 450nm, not 460, 460 is actually the standard blue and 480 you mentioned is getting towards cyan, i would also suggest not using green or red, warm white will already cover both of those spectrums and red light seems to be bad for coral growth since red doesnt naturally penetrate water much, but you will be creating a huge spike where in most reef lighting that we see now there is little red, i guess you could go with the fiji purple? that has a red component.

i would also tone the regular blues 460nm down, maybe to 4-1 for rb and blues, they are very windexy but still great to use to fill in the higher wavelengths of blues because they fluorese different colors better than the royal blues.

also just to be a pain, there is a new xm-l that cutter is carrying now that is a little more effecient than the xp-g, the real kicker is that it can be driven to 3000ma and produces something like half the heat of earlier leds. the downside is that they are very blue, something in the 7000k range.

nanotuners also carries a 420nm led now but the current capabilities are very low and the light is not very visible, probably not a good option at the moment but its a slight posibility .

if you were insistant on the 48 leds, i would probably go with something like

13-14 xm-l (driver1)
4 xm-l 4 neutral white 4 warm white (driver 2)
12 royal blue (driver3)
8 royal blue 4 "cool"(regular) blue (driver 4)

i imagine this would be a fairly low color temp at max power which you should probably never need to hit especially with 4 T5s

as for the warm whites, supposedly the xp-g warm whites will be coming out with a new line that will have high cri like 80+ which would imply the new technology with the boost in effeciency in the red phosphors
 
zoomed.com
hikariusa.com
aqaimports.com
Store