LED Experimental Tank (to be started later) - Page 5 - DIY Aquarium Projects - Aquatic Plant Central

Go Back   Aquatic Plant Central > General Interest Forums > DIY Aquarium Projects

DIY Aquarium Projects For those that are handy or looking to save some money, discuss your DIY aquarium projects here.

LinkBack Thread Tools Display Modes
Old 01-19-2009, 12:53 PM   #41 (permalink)
Junior Member
Join Date: May 2005
Location: Colorado
Posts: 12
iTrader Ratings: 0
felis is a regular member
Default Re: LED Experimental Tank (to be started later)

A quick look at the datasheet shows that white LED have largest luminous output. They are also most expensive. You need to compare prices/lumens and see which color gives you the cheapest lumen.
felis is offline   Reply With Quote

Advertisement [Remove Advertisement]
Old 01-19-2009, 02:14 PM   #42 (permalink)
Junior Member
Join Date: Sep 2005
Posts: 16
iTrader Ratings: 0
mcjosh13 is a regular member
Default Re: LED Experimental Tank (to be started later)

Luminous flux is a measure of how bright a light LOOKS to the human eye which has a peak sensitivity around the green area of the spectrum. Photosynthetic organisms do not care about this measurement as they are largely insensitive to green light and grow in response to blue and red light.
mcjosh13 is offline   Reply With Quote
Old 01-23-2009, 08:41 PM   #43 (permalink)
Join Date: Dec 2008
Posts: 48
iTrader Ratings: 0
Cory Keeper is a regular member
Default Re: LED Experimental Tank (to be started later)

Sorry I've been gone, But regarding how many LEDs you need for a tank, a good rule of thumb I've devised is for low-medium light, 1 LED per 24 square inches of surface area of the tank. I.E. my 29g has 15 LEDs over with the top having a surface area of 360 square inches. Higher light would be in the 1:18^2" or for insane light 1:12^2" ratio.

Do not attempt to use Lumens as a rule of thumb, as LEDs get better at emitting light, keep in mind PAR levels may or may not jump.
Cory Keeper is offline   Reply With Quote
Old 01-26-2009, 11:23 AM   #44 (permalink)
Join Date: Jan 2009
Posts: 1
iTrader Ratings: 0
ElectricFishMan is a regular member
Default Re: LED Experimental Tank (to be started later)

I looked at doing this and decided I'd better wait until two things happened: I had more free time, and high intensity tricolor LED's came down in price a little bit more.

I'll post my concept here so someone can take some inspiration from it:

I wanted to develop a programatically dimmable fixture. There are two ways to dim an LED, either reduce the current from your driver (you'll get mixed results as the forward voltage curve for LED's vary greatly even if they're off the same wafer) or use a Pulse Width Modulated (PWM) power supply. My design opts for the second, and more easily realizable PWM setup. PWM relies on the fact that the human eye (and I'm assuming fish are similar here, any help on this would be appreciated) doesn't detect light changes at a rate of higher than about 30 cycles per second. Instead the eye averages it. That's the same principal that your television takes advantage of (try snapping a photo of your tv screen with a regular 35mm camera). By adjusting the amount of on and off time during each cycle you can adjust the overall light intensity.

I like the idea of using tri-color LED's, especially if PWM is being used. With tri-color LED's you can not only make your tank glow white, but any color you want by varying the intensity of the Red, Blue, and Green LED's. Can anyone say mood lighting?

And for the real over-achievers out there: PWM can be easily set up with a $3 microcontroller (I like the PIC micro's personally). They also provide code for maintaining a real time clock and doing basic math calculations. If you really want to have some fun, it's feasible to create a daylight calculation and vary the intensity to mirror a normal sunrise, day, sunset, night cycle for an arbitrary latitude. For the night cycle PWM can turn off the lights, or better yet, dim their intensity according to the lunar cycle.
ElectricFishMan is offline   Reply With Quote
Old 01-27-2009, 08:21 AM   #45 (permalink)
Join Date: Nov 2007
Location: Seattle
Posts: 98
iTrader Ratings: 3
iTrader Positive Rating: 100%
longbowaj is a regular member
Default Re: LED Experimental Tank (to be started later)

I'm torn between adding a lot to this conversation or keeping quite. I have worked on the mood lighting systems for Boeing (777, 787, 747-. I know how they work fairly well and I will say they are very complicated, but most of the ideas could easily be applied here. ElectricFishMan has the right idea with the PWM but like I said this could get very complicated fast depending on how detailed you start writing your algorithms.

I would like to add one thing right away though. Don't believe much of what you see on LED spec sheets. All the suppliers are notorious for lying about what the outputs are. This is going to make it very difficult to really be scientific with using LEDs in aquarium lighting unless you have a light lab to take your own measurements.
longbowaj is offline   Reply With Quote
Old 02-09-2009, 07:55 PM   #46 (permalink)
Junior Member
Join Date: Oct 2006
Posts: 5
iTrader Ratings: 0
Bizarroterl is a regular member
Default Re: LED Experimental Tank (to be started later)

For a algae scrubber you could use this:
or you could use as part of your overhead lighting (plant specific).

For general tank lighting this looks good:

More DIY is this one:
It's less expensive, claims less watts, but has more LEDs.

None of the above are dimmable.

Homegrownlights.com has a 50watt panel that is DC and would require a power supply. It is dimmable with the right power supply and control.
Bizarroterl is offline   Reply With Quote
Old 02-10-2009, 03:57 AM   #47 (permalink)
Senior Member
Join Date: May 2007
Location: Lincoln UK
Posts: 398
iTrader Ratings: 0
Supercoley1 is a regular member
Default Re: LED Experimental Tank (to be started later)

Hello. I've just read through this and see there a few doubters out there and a few people giving advice that seems strange to me. I would however add that I am no electrician so all these people 'know' more than me Any advice given below is what I was told whilst making mine. I had never soldered anything let alone understand electrics so I was guided through mine by several interested electrical people

I have just made a luminaire using 15 x 3W 'Luxeons'

Firstly the ebay seller selling 'white' Luxeons is selling 5500K LEDs they seem to be copies as some have Lumileds branded on them whereas others don't however I can say they are in the main pretty good. 5 for 10 is worth it even if only 2 thirds are a good match lightwise.

Secondly I wouldn't use lenses/collimators on these. They are lambertian dies and therefore give a good light spread at a 120 angle. When used with lenses/ collimators they give a point focus which doesn't look that good. I assume this is why the 'retail' luminaires are slated by reef keepers in comparison to MH!!!

'Batwing' dies would be good for our setups without collimators/lenses also but not side emitters for obvious reasons.

I'm not a fan of the Lumens per square inch calculations nor the WPG rule. You can't really use this on LEDs anyway. Much better is to see what is out there see what their setups look like gauge their opinions and observations on growth.

These LEDS (3W) get awesomely hot. They need proper heatsinking or they will fry. I also have fans on mine to help alleviate the heat.

You must use current controllers and not resistors and not just assume a 12V regulated supply is supplying 12V. The V allowances of these is very limited (3.4-3.7 or similar) and therefore any surge etc will fry them. I was also told to always wire high power LEDs in series.

From my observations the initial posters comments on 18W high power LED being equal to 36W of PC are not far off the mark. I would go aas far as to say that the plants under my setup (running at 42W) is growing much much faster than it was under 48W of T5HO. I would say they are probably 1.25 - 1.5x better than the T5HO was!!!! CFs are nowhere near T5HO IMO (and experience!!!)

So a little look at my setup. I can answer some questions as long as they are in non electrician's English

The setup as I said before is using 15 x 'Luxeon' 3W 5500K stars. There are 5 series of 3. Each is run by it's own current controller. Each series is powered by a seperate '12V 700ma+ regulated adaptor'.

This can be done by using 1 power source of course and computer with electrical knowledge but I wanted a sunrise/sunset and with my abilities I went for this option as then I can put each adaptor into a timer.

Therefore I have in effect got 5 seperate LED setups within the 1 luminaire.

Each Star has a computer memory chip stuck to its rear with thermal tape. These are too small to be used as heatsinking on their own so the LED/heatsink combo is then attached to cut up pieces of an old hollow reflector I had handy. Not for reflection more to use the extra metal for drawing more heat out.

Each LED unit complete with heatsinking is then attached to a board and wired in series from front to back (5 rows from left to right.) then each series is attached to it's own current controller. The particular current controllers I got says they are for 7V to 24V. It uses 1.5V for itself leaving 10.5V for the series (3.5V each.) There are then 4 x 50mm PC fans blowing in from the left and 4 more 'sucking' on the right. These are all underpowered running from a single 12V 300ma source and therefore are quiet (almost silent unlike our PCs) Then the whole unit has a pane of glass sealing the bottom so the airflow doesn't go downward and travel from entry fans to exit fans They also run 15 minutes and 15 minutes after any lights. the former to dry any little moisture that may be in the unit and the latter to cool down the LEDs at the end.

These LEDs are 3.7max and therefore I am very slightly underpowering them at 3.5V (10.5V/3 Leds) which by a sloppy calculation that will not be accurate I say they are each running at 2.8W and not 3W. This of course will not be exact but is good enough for my purposes.

So there we have it 5 series. Each series running at '8.4W'. Total '42W' over a 33USG tank. Easily brighter than 48W T5HO. The T5HO was at 4-5" from the water surface. I have had to raise these LEDs up to 13" above the water surface so they don't wash the colour out of the top of the tank (meaning you can't see the colour in the top third due to the intensity of light near the surface.)

I then made a simple single series for a 10 Litre Nano using the same LEDs from the same supplier, same current controller but used the recommended heatinks and 30 collimator/lenses.

This picture show an individual LED with the mini heatsink on the rear, the piece of reflector used as extra heatsinking and then the whole lot screwed down to the board.

Here is the whole board. looks messy but this tape came off after siliconing all the wires down to stop movement when it was turned upside down

Here is the board from above showing in red 1 series:

The board complete with Cold Cathode 'moonlight' all lit up:

The luminaire ready to hang up.

Gotta go and moderate somewhere now so let me know if you want me to carry on and I will try and get some more pics on tonight for you to see.


Last edited by Supercoley1; 02-10-2009 at 04:06 AM..
Supercoley1 is offline   Reply With Quote
Old 02-10-2009, 08:55 AM   #48 (permalink)
Join Date: Jan 2009
Posts: 34
iTrader Ratings: 0
andyh is a regular member
Default Re: LED Experimental Tank (to be started later)

Your setup looks great!

It appears to me that you have 15 X Luxeon 3W, running at 3.5V and 700ma. My calculations say that you're running the Luxeons at 2.45W - a little lower than your calculations, but maybe I'm missing something. If you have 15 of them, my calculations say that you have 36.75W of LED'a over a 33g tank, right? Since you feel PC and LED wattages are not comparable (and I think I agree from my early tests), let's call your 36.75W of light 36.75 "LEDW". So, you have 36.75 LEDW over 33g, or 1.11 LEDW/g. If I had the dimensions of your tank, I'd do a LEDW/SqIn calculation also - for those that like that surface area calculation better.

I'm going to be running 60 X Cree XR-E, and running them at 3.7V and 1000ma. That comes out to 3.7 LEDW, which is as far as they can be pushed while staying in the manufacturer's specifications. I will be running that over a 135g tank, which means that I'll be at 222 LEDW over 135g, or 1.64 LEDW/g. Since my 135g tank is 72 X 18 X 24, I have 1296 sq in of surface area. That makes .17 LEDW/SqIn of surface area. I know that Lumens and PAR and all that are as important or more important than the above, but when you're trying to figure out a rough estimate of how many LED's you'd need to make your planted tank work, this seems like a good rule-of-thumb for estimation.

What I'm getting at is this question: "why don't we start (maybe in yet another thread) a list of plants that will grow well at certain LEDW/g or LEDW/SqIn?" That way a person considering LED lighting would know how many LED's they're going to need to successfully grow the plants they want to grow. I'll help to populate the list (I'm still prototyping as most of my LED's are still in the mail from Hong Kong).

Maybe the right calculation is "LED Lumens per gallon" (LEDL/g) or "LED Lumens per Sq In of surface area" (LEDL/SqIn) - time will tell.

Opinions? (all experienced opinions welcome)

Now for a personal experience question: CoryKeeper (who seems to have experience) feels that my 60 X 3W Crees will supply a "good medium amount of light" over 135g. That's 48% more LEDW/g than you have (although that's probably a bit high as efficiency falls off the harder you drive the LED's), and you say that your plants are thriving. He seems to favor a Surface Area calculation. What is your opinion (from your experiences) about 60 X 3.7W Crees over 135g and 1296 SqIn of surface area? I'm looking for enough light to grow my favorite Ludwigia Cuba, which doesn't seem to keep it's red coloration without plenty of light and CO2. See why I'd like to know in advance if I'm buying enough LED's? Let's help others that would like to estimate their costs and lighting by giving them a way to figure out what they need with a bit of confidence!

PS: I agree with the "awesomely hot" comment about LED's at (and above) the 3W line. I'm experimenting with water-cooling and tank heating (see other threads) with the heat. They may not be as hot as MH lights, but they are hot enough to really need to work on it.
andyh is offline   Reply With Quote
Old 02-10-2009, 09:29 AM   #49 (permalink)
Senior Member
Join Date: May 2007
Location: Lincoln UK
Posts: 398
iTrader Ratings: 0
Supercoley1 is a regular member
Default Re: LED Experimental Tank (to be started later)

I will put the pics up later but a quick(ish) reply to your post

I have no idea what wattage they run on. You can do the calc for me so I know

These are supposed tobe 3.4V to 3.7V and I am running them at 3.5V.

I think they are growing 25% faster than the previous fluoros (which were 30W T5HO, 18W T8.) this is judged by my needle ferns which I had to prune to see light on the substrate before. I am now pruning weekly and heavily at that.

Therefore if you say I am running at 1.1WPG on LED and I was on 1.45WPG of fluoro yet I have more growth then that seems to answer one question although of course you have to take my word on the growth

I did a test with some Rotala Macrandra in the tank under the T5HO alone a few months back and it kept a peachy red colour under 0.9WPG T5HO. Therefore I would assume I could grow it under these LEDs. I will test again when I rescape and have some room for the test again:

This will stir up some haters for me but I don't like PC/CF over tanks. I prefer to spread the light than have a huge intensity stuck in one area. I would take T8s over CF for this reason. T5HO are much much brighter than T8s. LEDs are proving to be even higher

the tank is a standard 125Ltr Tall. I think it is 800 x 345 x 450 (mm)

I'm not really in favour of another rule just as much as I hate the WPG rule. Lighting causes more disagreement than any other subject to the point of confrontation sometimes as people try and use science and power to equate to plants and nobody knows for sure how much light even in PAR certain plants need. It would just be another WPG rule for people who like the Lumens per sq inch calcs and it all gets a little bit silly really.

I would suggest that we know that we can dose EI and ferts are then non limited. We know we can reach 30ppm and maintaining it means that CO2 is non limited. With me believing that CO2 is the key to all these 'highlight' plants growing then I am assuming there is a level of light where you can grow anything and if the ferts and CO2 are maintained at a non limited level then success.

Now my assumption is not based on science, it is not based on proven fact and I have no PAR meter nor scientific knowledge in the least to say that my theories are true so please take them as that. They are theories from my own observations and assumptions.

I think that if you have 2 T5HO (say 1.5-2WPG) over an 'average' tank (under 2ft) and put them a third from the front and a third from the back then they will grow plants better than 3 T5HO all grouped together in the centre even though the 3 is more 'lumens' than the 2!! Good spread is what I want to achieve.

Once that good spread is achieved and the light is good enough (1.5 - 2WPG T5HO) then you have a decent level of light to achieve your HC carpet or luminescent red plants as long as you keep the CO2 high AND get it down to your carpet and red plants. Adding more light may get plants to grow faster but it won't make them redder or carpet HC to a mm of the substrate It also means you have to work harder to maintain high CO2!!!

This is why I hate all these rules. Mainly because you create a rule and someone disagrees with it and then I come in and contradict your rule and then they disagree with me and then noone reading knows what to calculate anything on. lol

Anyway I think if you are after high light over a large tank like yours I would be aiming for in the region of 200W of LED maximum dependent on how high you are suspending it (I am however judging this on observations of my LEDs/drivers/input voltages etc) which may differ from yours

The Lumens of these Luxeon III 3W lambertians 5500K @ 700ma are apparently 80 lumens if that is any help to you Differs per colour die though remember that. The blues are 30lumens


Last edited by Supercoley1; 02-10-2009 at 11:14 AM..
Supercoley1 is offline   Reply With Quote
Old 02-10-2009, 10:47 AM   #50 (permalink)
Join Date: Jan 2009
Posts: 34
iTrader Ratings: 0
andyh is a regular member
Default Re: LED Experimental Tank (to be started later)

While it's not exactly right, watts and "va" are used interchangeably at reasonably low power - so a rough "wattage" (actually "va") for your LED's is 3.5V X 700ma (which is .7 amps), or 2.45 watts.
At 3.7V and 1000ma, my Cree's are specified at 228 lumen max (Q5 bin).

I agree with you about "too many rules", but I see time and again on this forum and others that people want to consider LED's, but just can't get an estimate as to how many LED's and expense they would be committing to. I'm not looking for another "rule" - I'm looking for an estimation method that says "if I want to grow this plant in my 29/55/90/135/200g tank and I have/don't have CO2, other people have succeeded at ??? watts of LED's". I have seen people think that a half dozen 5mm junk LED's at 20 cents each are going to do the job, and others that believe that the $4,000 Solaris units aren't enough.

I'd just like to help others get a reasonable quick guess at what took me countless hours over 5 months to figure out.

Last edited by andyh; 02-10-2009 at 10:50 AM.. Reason: forgot the Lumens on my Cree LED's
andyh is offline   Reply With Quote

Aquatic Plant Central > General Interest Forums > DIY Aquarium Projects > LED Experimental Tank (to be started later)

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On

All times are GMT -8. The time now is 10:11 AM.

Powered by vBulletin® Version 3.8.8
Copyright ©2000 - 2020, vBulletin Solutions, Inc.

Search Engine Friendly URLs by vBSEO 3.6.1