PRIMARY LED lit Custom glass tank

[wetWolger]

Humm, I have been doing a lot of research into LED arrays for ummmm.....growing _plants_. I have done a lot of reading of NASA pattents and stuff which gives a lot of great info. Here is one in particular I find very intresting:
http://patft.uspto.gov/netacgi/nph-...0&s1=6921182.WKU.&OS=PN/6921182&RS=PN/6921182
It deals with terestrial plants mostly so the red blue ratio is worthless for this project. It also talks about how they wired the LED's. They found you get the best results if the leds were mixed into each other instead of having a row of one kind and then a row of another kind. But if you wire diffrent LEDS in the same line it can cause problems as stated in the link. Not sure if you did that. Also your LUX reading are almost worthless because as you may or may not know LEDS are very very narow banded. The white leds are accually white because they are flashing back and fourth from two diffrent bands to trick the human eye into thinking it is white. So you should try taking a reading with all the white leds off. Infact if you took a reading with just one type of LED running at a time you may be able to get a pretty decent band spectrum since you would know what band they are releasing. Based on reading everything you have said IMO you don't have enough light for sps and clams. You would be lucky if you can keep softies.

Also for your burnouts LEDS give off very very little heat. And they can withstand very very intense heat. I have a led I installed in my car that probobaly reaches close to 100F and runs constantly. It has not burned out and has been in use well over a year. My guess is they are burning out due to your "overdrive period".

Also you say your output was twice as high two months ago, leds lifetime are like 300,000hours. And after about 100,000 hours there intensity decreases by 50%. So I would think you should easily get 30,000. So my guess is this is also due to your overdrive.

Hope that info helps.

 

[kgolem]

<a href=showthread.php?s=&postid=7029278#post7029278 target=_blank>Originally posted</a> by zachtos
If any more problems like this come up, I will scrap the entire project. It's not worth the daily stress anymore, this is not relaxing in the least, especially since I will be losing more and more money. I lose 2-3 white LEDs per week, I can replace every white LED but that will take me about 8hours of work and cost around $70. screw this. very very close to getting MH and declaring LEDs are feasible IF you can get a good manufacturer.

All corals are still alive and growing slowly, corraline is spreading along the foam at a slow but steady rate. I have not purchased more corals or own any fish still due to the LEDs constant failures, I've been hesitant to add anything. I miss when the output was twice as high 2 months ago. I'm afraid to replace all the LEDs and waste more money.

Zachtos;

I have been reading this thread for a few weeks now and following along with great interest. I work with several electronics designers like myself, and technology school grads. Today we were discussing your project, we agree on a couple things that may help you out.

You purchased your LED's at a remarkable discount from retail in North America. I suspect that your LEDs were obtained at a similiar discount from the manufacturing facility in the far east. Perhaps they are from a production run that failed some kind of quality control test and sold out the back door of the factory for pennies on the dollar. This would explain your high failure rate.

Looking at your circuits, there is really nothing wrong with your design and concepts.

You should continue with your project, however, I might suggest a couple things, and please understand that I am being constructive here, no ill will intended.

More attention should be paid to prototyping your project. Instead of building the entire array and firing it up (I know that you tested your circuits), it might have been prudent to build a single string of LEDs and test over the long term, not less than a period of 3 weeks of continuous on time, with periodic measurements of current draw and light intensity, temperatures etc. Use a statisically significant percentage of your LEDs for the protoype, 5 to 10 percent would be enough, to test your designs and the integrity of the components.

Try different circuit designs, different LED products. A good thing to try would be to design around a smaller supply voltage, or a different type of supply, try an ATX supply from a used computer instead of a wall adapter, as a suggestion.

It sounds like you learned a great deal with this project. Don't be discouraged. You are breaking new ground for the rest of us here. The proper approach will minimize the frustrations and at the end when your array is working perfectly, the reward is so much better.

 

[melev]

It would be nice if interested people could send Zachtos 10 white LEDs each, but again the quality of the bulbs would come into question as their sources would be varied. But sending him bulbs to continue the project would save him a little money and not break the rules here on RC about seeking donations, as far as I know.

Sorry it hasn't worked out. I think getting the LEDs at a discount could be the problem as well. I bought PC bulbs that were on clearance a few years ago, and they barely lasted 6 months compared to normally-priced bulbs that were good for 12.

 

[zachtos]

The only successful part of my project thusfar is the fuge light and moonlights. The moonlighting looks fantastic and my macro algae grows very fast and produces no heat w/ the lighting array. I dont get any burnouts on my UV/blues/reds and they look pretty sweet... in the morning my tank glows neon pink. I have proven that corraline and coral can grow and survive under LEDs, but the growth rate is not astounding as of yet. I can only keep softies in this prototype tank because small water volume size and inability to keep chemical warfare under control. I'll think about ordering replacement whites and redo all the whites since so many people are following the thread. I'll try and get a photo up again since I'll be losing access to my friend's gf's hi-megapixel camera soon and be back to my old crapo webcam.

No need to send me LEDs, I don't want to have to try and determine the required resistor size again. That is a real PITA. If I order again, it will be the same manufacturer/model. I'll just have to hope that they wont burn out since I wont be overdriving them this time as much. I think they are rated max of 30mA and the max I pump through them now is around 21-22mA... havent tested in a while when hot though. I will this weekend when I replace more burnouts.

 

[perpetual98]

I wonder if those besthongkong MR16 LED arrays would work in your standard MR16 track lighting jobbers? I've got a bunch of those at home that I'm not using and they each of a 12V transformer built in. Would be great on a frag tank, or using supplemental lighting on specific corals in a grow-out tank, probably not for general tank lighting.

 

[zachtos]

full tank shot after nearly 2 months of LED light only

the fan I installed to help remove heat

small corraline growth can be seen on the yellow parts of the greatstuff foam easily

I double this hanful of macro each week with the pink LED glow you see in the background (reds/blues/UVs)

I'm still thinking of rewiring all my whites to repair all the damaged overdriven ones. I may order them this weekend.

 

[leeweber85]

glad to hear you're not giving up!

 

[thriceanangel]

Keep going! I am very interested in this. I think that LED will be the way to go. Edison didn't give up after the first try. Take a break if you need to but you are providing ground breaking reasearch here! Keep going. It could be the cheapest, most efficient lighting available, and if you can get it together you can have a patentable fixture! $$$! Keep it going please!!!

 

[zachtos]

I have replacement LEDs to re-wire my entire array when I have time to give up a day. I just am not motivated enought to do it still... I do have a steady growth of corraline in my tank and my Green star polyps have actually replicated a bit. Everything else is not really showing growth but the tank is just reaching a mature stable point now. I think once I renew all the damaged LEDs it may do better.

 

[Aquaduck]

I believe your problem is that you did not "bin" your LED's by their forward voltage drop. So what happens is the one with lowest Vf starts to take more current and burns out. Next time you make a string, check the Vf and separate them into groups with common Vf measurements. Use only those LED's that have the same Vf in the same string. Otherwise you need to wire them in series parrallel combination so they all share the current equally and I think that would be a lot of work for you.

BTW I commend you on doing this project. One day I will build my own fixture with Luxeon LED's and use a DMX controlled driver to adjust the brightness and colors independently.

 

[Aquaduck]

Patrick,

I can't make it to the show. Would you be able to post pictures of your booth and LED fixtures? The company I work for is releasing a dimmer/driver for Luxeon LED's. IRIS LED Driver I did the mechanical design and PCB layouts. I also designed and built some PAR38 LED fixtures for an upcoming lighting show.

 

[easttn]

Patrick,

Are you using 1W, 3W, or 5W LEDs? I'm a manufacturing guy, if you need any assistance please PM. My engineering team has built Hayward and Pentair pool and spa lighting products for over a year. We also developed a patented algorithm that will deliver a specific luminance and chrominance geared to Luxeon's binning parameters. Our robotics engineers have developed a soldering system that will not stress the lamp. Are you using a metal clad PCB? Lots more questions.

If you understand the specifics of this post, I can help you.

Charles

 

[hllywd]

<a href=showthread.php?s=&postid=6350928#post6350928 target=_blank>Originally posted</a> by zachtos
2 tsp of human urine, 24 hours later I have 2ppm ammonia. LEDs are still not here. Entire project on hold waiting.

I guess now I've heard everything.

 

[jnarowe]

What, you haven't heard of the urine & vodka cycle??? :lol: Where have you been?

 

[hllywd]

I've heard of dosing vodka.... but pi__ing in my tank??? Sorry. Personally I think both ideas were born out of too many Cosmos...


BTW I like the LED Idea and was subscribed to another lengthy thead on the topic I haven't heard from in a long time. I'll see if I can find it. That guy had some really nice pics of some zoas that looked great under the LEDs.

 

[hllywd]

<a href=showthread.php?s=&postid=6294893#post6294893 target=_blank>Originally posted</a> by tekknoschtev
lscaquarium, they make type 1 and 2 with the anti-mold and anti-fungus agents. This whole point has been debated to no end, and the "not for aquarium use" is a fall back or a "cover our butts" statement made by GE and many silicone manufacturers. There are different strengths of silicone, but GE Type 2 doors and windows is perfectly fine.

Before it gets out of hand though - I reccomend that people do some of their own research (and yes, they may mean waiting for the search here, or working around it with google) and decide for themselves which silicone is appropriate for their application. When I asked a while ago, I was told GE Type II Doors and Windows because Type I had the anti-mold stuff in it. Later I was scolded for using Type II and told my tank would fail and I'd have a pool of water in the living room. While its only been about 6 months - the 150 is still holding fine, all of the baffles in the sump are doing well, and the tank is rock solid with no leaks. While 6 months isnt proof in and of itself, I feel as though its safe and sturdy enough to pour my paycheck into livestock for the tank

My question is if the "not for aquariums" statement is a CYA (completely separate form the "limits of liability" statement on the label), why Does the GE II have it but not the GE I. In the past I thought I used the GEII and it had an "ok for food contact" statement and no aquarium warning. I can't find that one anymore and the GEI is silent on either point. It may have been a DAP product that I remember the "food use" statement from. At any rate if it says "not for aquariums" and it's counterpart doesn't for the same price, why would you take the chance? To be defiant because you assume it's a CYA? Doesn't make a lot of sense to me and I feel is bad advice.

 

[hllywd]

http://www.nano-reef.com/forums/index.php?showtopic=57855&hl=led

 

[zachtos]

Yes I followed that old LED thread and there was not much followup. that thread among a few others at nano-reef.com caused me to start my project. i am still trying to find a day to devote to rewiring my whites. as for the pee in the tank, I just wanted to kick start the amonia cycle, it worked and it was free, so lets not get bogged down on that topic. The GE2 stuff is fine as long as you dont use the anti mildew kitchen kind. the great stuff is also fine. I have received no unknown deaths, but my kenyas seem stunted. My great stuff foam is now 25% covered w/ corraline and starting to get feather dusters. It's taking time, but I think it will speed up when I renew all the damage whites and start paying attention to the calcium more, which I have not tested in months. The alkilinity stays around 10dkh and salinity 1.025, those are the only tests I bother with now. My nitrates etc are all 0ppm usually. And my macro is still growing w/ my red/blue/uv LEDs at a steady rate proportional to my tanks bioload. I still have not bought the green clown goby I want nor any new corals, again... waiting till I get around to rewiring.

 

[zachtos]

I finally forced myself to renew my entire array. It is the same as before but I renewed all of the whites and UVs. The reds/blues are all still strong/undamaged apparently. It took nearly 7 hours to remove/rewire/resolder the array. But I am hopeful that I have a stronger design now. The white LEDs are from besthongkong.com again and cost nearly $90 for 300. they are 40,000mcd and tend to operate at 3.3V at 20mA.

*overview of design improvements*
-each strand has 75% more resistance then original design, which means each strand will pull less current, and should generate less heat, which means it will damage the LED less while putting out reduced lighting. But this is a good thing considering my light intensity declined by a factor of 5 in 5 months.

*detailed design improvements*
I bumped up my resistors to 160ohm - 0.25watt. I have strings of 8white:2blue w/ 160ohm resistor w/ 36VDC across each. The LEDs are rated for 30mA. I measured around 3.4 to 3.8V as the temperature increased across each resistor which means 22mA on startup then 24mA at operating temperature. I dont anticipate the LEDs dropping to less then 3.15 to 3.1V each MIN which would mean 28-31mA which is the max of the LED. I'm hoping they will stay at 3.2V each MIN so it will pull 25mA each. The UVs now operate around 19mA each or less now so they wont wear out so fast this time.

--------------------
Results
--------------------
The lighting at the surface is now 21,000 to 26,000 LUX as opposed to 2000-5000 LUX after only 5 months of operation. This is a tremendous visual improvement as at the end of the lifespan, the tank looked very shadowy/blue. The tank is now very very white. Nearly 25 white LEDs failed after the initial over current damage of the first 2 weeks of operation. The LEDs continued to fail even after I added more resistance, likely due to initial damages and incredibly reduced lifespan.

No photos, I do not have access to a nice camera anymore, but will try and get some kind of photo despite poor quality up soon.

 

[kgolem]

Congrats, it looks like you are on the right track this time. I am looking forward to pics.