Masters of LED lighting, please help me build a light.

[Undertheradar]

PFO has just come out with a nice unit, but I cant help but think it wouldnt be that hard to make something similar out of high intensity LEDs. I would like to make something along the lines of the PFO unit...

 

http://www.pfolighting.com/Aquarium-LED-Lighting.aspx

 

but I dont see why this couldnt be done for much less. I would like to use white LEDs, some 420nm (actinic), and blue (450nm) for a good blend.

 

Where do I begin? I would like to make something that rivals a 250watt halide.

[Noycey]

i've been messing around and researching led's and so far i think to rival a 250w MH would cost way more than buying the MH but you would have the advantage of longer running life (not sure if this would offset the initial cost).

 

Not possessing the relevant qualifications to understand if led's could provide the right spectrum of light my current thinking is i need to somehow measure the light frequency range given off by a mix of led's at different intensities.

 

my 0.02

[Undertheradar]

the spectrums for most of the LED's I would consider are already tested and shown. The 'superbright white' LEDs have their spectral curves shown in many places...

[TheGent]

I have seen the spectral curves of the Nichia brand white LEDs, but I have not seen how this coorolates to the tempature of the white light. In fact, from what I can tell from my internet research is that the measurement for tempature is completely different from the spectral curve. I know that the side of some lights show the tempature graph, that looks just like the spectral graph, on the box, but that is the only place I have seen tempature represented as such.

 

LEDs are great, especially in these nanos. I have seen some of the projects here and it's really tempting to go ahead and do one myself, but it is a hell of a lot of wiring, and it requires quite a bit of power.

 

If your going to do it yourself, please pay attention to the maximum forward current on the manufactures data sheet. I've seen projects on here that use LEDs, but they do not use any resistance. Doing that, you won't know what your spectral curve looks like and you'll burn out the LED eventually. Point of reference: a green LED will look yellow if there is no current limiting resistor.

 

Another thing to look into is instead of doing all the wiring by hand, get a PCB manufactured. I know there is at least one place on the web that can do low run PCB turns at a minimum cost. This wll make everything easier for you in the long run.

 

Remeber that the LED lighting is direct. There won't be much spill over, so you'll need to run them the lenght of the tank. There are also some tricks you can pull with plastic covers that will avoid hotspots, but at the same time, it will cut down on the intensity.

 

I'm an EE and most of the stuff I work on is for human interface devices. I have a lot of experience in LED's, LCDs, ect. So let me know if you need a hand.

[mattgreene22]

TheGent - Where can you get a PCB manufactured cheaply? And do you have to do more than one for them to take the order??

 

cheers

[soundman]

Sorry, but I know nothing about LED's or electrical matters. I am interested though. I've been wondering if LED's could be substituted for the PC's in an Aquapod (or Nano-cube) and still deliver the intense lighting needed for corals without the heat of the PC's & ballasts. The link you posted above for the PFO light makes me wonder about the heat issue though because of this statement:

 

"The LED’s will have premature failure if they get too hot for an extended period of time. Ensure that there is good air circulation around hood."

 

Does anyone have any idea how the heat put off of LED's compares to PC's or even MH's with compareable light output? If the LED's can live up to the 50,000 hr to 100.000 hr life expectancy while maintaining the proper light spectrum their entire life, that would be great. 50,000 hrs @ 12 hrs/day = 11.4 years of not having to replace PC or MH bulbs.

[neanderthalman]

I've seen projects on here that use LEDs, but they do not use any resistance. Doing that, you won't know what your spectral curve looks like and you'll burn out the LED eventually. Point of reference: a green LED will look yellow if there is no current limiting resistor.

 

I think you're referring to my LED project. You're close, but not quite right. A green LED without a resistor will not look yellow. A green LED without a resistor and a voltage supply higher than the LED's max forward voltage will look yellow. So long as you match your voltage supply voltage to the LED voltage, then there's no need for a resistor, and the LED will operate normally. You're an EE, so you know all of this - I'm just tossing this out for those who do not have an engineering background.

 

Again, my LED project is a case in point. I've matched the LEDs to the power supply, and a year later, it's still running perfectly. There has been no negative reactions by corals or algae growth to indicate a shift in spectrum, and I can count the number of LEDs that I've lost on a single hand.

 

 

Another thing to look into is instead of doing all the wiring by hand, get a PCB manufactured. I know there is at least one place on the web that can do low run PCB turns at a minimum cost. This wll make everything easier for you in the long run.

 

Remeber that the LED lighting is direct. There won't be much spill over, so you'll need to run them the lenght of the tank. There are also some tricks you can pull with plastic covers that will avoid hotspots, but at the same time, it will cut down on the intensity.

 

The PCB is a good idea, you can get the supplies to make your own PCB's for around thirty bucks and it's fairly easy to do. I've got the link saved on my home computer, I'll post it later. There are lenses that you can get that will spread the LED's illumination out further, and you can also get LEDs with a wider veiwing angle. The mcd measurement of such LEDs will be lower, although they have just as much lumen output. This is because the measurement of mcd takes the viewing angle into account, and a tighter veiwing angle will result in a higher mcd from what is otherwise an identical LED.

 

The only issue with the PCB is that the cooling of the LEDs is not as efficient. Most of the heat will be removed from the LEDs via the leads, and by leaving them fairly long and soldered together in open air, it becomes something of a heatsink. A PCB does not create this array of leads for air to be forced through, so more careful consideration of cooling should be given. A fan blowing directly on the backside of the array *should* be enough to draw the heat away effectively.

 

I did a few calculations a while back, and LEDs were not cost effective or energy efficient on aquariums larger than a 5.5gal because of the tight focus. If you're looking to set up a pico UTR, then LEDs work quite well. If not, then MH is probably a better choice. What size of a tank are we talking about here?

[TheGent]

www.pcbexpress.com is one place I've used. It's cheap for PCBs, and your minimum is 2. That place will change the price depending on the size of the board, so you could be in for a world of hurt if your not careful.

 

pcbfabexpress.com is another place that I have not tried but plan on using in the future. Thier prices seem better, but the minimum is 5 pcs.

 

When thinking about doing lighting like this, I picture a series of modules that you can plug in when you need to replace a part. So, maybe you have 4 squares of leds that connect to a power control board. If a couple of the leds on one square go out, you just plug in a replacement.

 

With that in mind, the power board could do cool things like moon cycles, vary the day time lighting and produce cloud effects (not that its needed, just cool).

 

I dunno, I've been kicking around the idea for a while. Maybe I can put together a schematic / layout for fun and see where it goes.

[Carinya]

FWIW, I was talking to a reef club member & we were discussing the PFO fixture versus my own unsatisfying attempt at LEDS and the fellow reefer said he had talked extensively with PFO about their bulbs. Apparently PFO is using some special, different kind of LEDs that cost them $4-$5 each (if I'm remembering correctly). That's why the fixture is so expensive.

 

I'm not sure if this is 1) true or 2) plausible since I don't know much about LEDs and I didn't have the discussion w/PFO. So, YMMV.

[neanderthalman]

It looks like they are using a luxeon or other ultra-high power LEDs, not your typical 5mm high-intensity LEDs that we've used for our projects C. At 4-5 bucks a pop, I'm seeing between $200 and $250 just in LEDs on that PFO light strip. How much are these things supposed to sell for?

[Undertheradar]

http://www.reefcentral.com/forums/showthre...ef&pagenumber=2

 

Im hahnmeister over at RC BTW. Above is the thread where PFO first broke the news to us. I was busy responding to some reefers who had never seen LED lights on reefs by showing some of what I have seen here...and then PFO goes and announces their invention.

 

"As far as shimmer effect. There is some shimmer effect. Not as much as Metal Halides but definately more than fluorescent. We are using 25 3 watt LEDs in a array of about 6" square per foot of hood." -PFO

 

From what I remember when I last saw the pricing, the single unit with the controller (you need the controller) was about $600. They are using only 75watts though, have a PAR that seems to be higher than a 250watt 20,000K...and you can control the output from 6500K to 20,000K. Im all for it, but the price is too high, and I just know that this is just the first of many mfg's that will come out with LED lighting systems in the near future.

 

I was thinking something powerful enough to make a pendant for a 20H, maybe a 40B at most.

 

For the mount, I was thinking a thick piece of sheet aluminum that I could drill holes in. Great heat transfer, great reflectivity, etc. I am in EE as well, so this would be a great learning experience and I have the staff at the university if I need their help. I dont mind spending a little $$ to get the project done...I really just need advice on where to shop for supplies and what parts I need.

[stevie]

I don't know much on LED's or electronics but still going to try a small tank. I will be using small hobby PCB and resistors. Been getting most off ebay so far. I also am set on getting a bigger tank in future maybe 40b but want to use T5 lighting only. Maybe you could do something like T5's with LED spotlights instead of MH?

[neanderthalman]

UTR, I'd look into using a higher power LED setup - something with luxeon LEDs. To do what you are trying to do with 5mm LEDs will have a hefty price tag and the labor will be astronomical. I don't know about the cost on the luxeons, however, but since you are building a larger fixture, issues like heat sinks are not an issue.

 

I'd be thinking to duplicate the PFO setup with a 25 LED array every foot or so. I'm working on an LED controller for my setup, but it and every other project of mine is on hold for god knows how long. Do you have access to a pic programmer at the university UTR? Perhaps in one of your laboratories? A simple PWM controller is fairly easy to program, and it can be used to give you a sunrise/sunset effect, as well as the shifting color spectrum and the random things like cloud simulation.

[Undertheradar]

Yes, I could integrate a pic programmer. I had not thought of that yet though. I am more interested in how I can get my hands on, and which LEDs I should try to get.

[neanderthalman]

I got mine through eBay, but I used 5mm LEDs, not luxeons. You could look on eBay, but be warned: It seems you get what you pay for as far as 5mm LEDs go on eBay, I imagine the luxeons (and imitators) are similar.

 

Have you had any luck looking through datasheets from lumileds (the makers of luxeons). I'd use them as the first source of information. Some of your EE professors might be able to help you source them cheaper than you could yourself. If nothing else, they might know which suppliers to use.

[Undertheradar]

I havent even started looking...thats what I would like you guys to help me with...Im no expert in LEDs. do you have a link for the lumileds information?

[pulse13]

I'm also planning on experimenting with Luxeons in aquarium lighting, I'll try to help out with your project if I can. I'd look into using the Luxeon K2 LEDs, they should be for sale from Lumileds through their Future Electronics store soon and possibly on Ebay. Checkout the datasheets for the Luxeon III and Luxeon K2 lines, also a company called Cree makes similar products.

 

Luxeon III: http://www.lumileds.com/pdfs/DS46.PDF

 

Luxeon K2: http://www.lumileds.com/pdfs/DS51.pdf

 

Cree Xlamp 3 7090: http://www.cree.com/products/pdf/XLamp3_7090.pdf

 

http://www.futureelectronics.com/

[CapHillReef]

If any of you design some kind of PCB for a pico LED light, I'd be very interested. I think there would be a market for something designed for the AGA 2.5G tank (11-13" long).

 

I'm a mechanical engineer, so I really can't help with the layout, but I'd love to help design some kind of enclosure. There are prototype houses that make metal fabricated parts these days with the same kind of quick turnaround and low quantities as a PCB. Or I could pull a favor from the guys downstairs at the machine shop at work to get one or two.

[Carinya]

Holy cats, the PFO order page has the 14" ones at $849!

 

Wow. Just total luck that the little nano I made that was supposed to be lit w/LED (and it didn't work out for me) happens to be 14" long.

 

But wow $849.

 

Has anyone seen these puppies in action?