48" LED Array - Up and Running...Mostly

[Ron99]

Well I now have most everything I need to build this so here goes. I'll start with pics of the parts. Here's the housing (an old Hamilton housing with acrylic splash shield) the array will go into along with my 3 heatsinks and two lengths of aluminum angle to mount the heatsinks in the housing.

http://hotimg23.fotki.com/a/70_70/254_23/LEDParts1-vi.jpg

And these are the 8 Meanwell ELN60-48P drivers for 80ish LEDs (each can drive up to 13 LEDs), 80 Cree XR-E emitters (half cool white, half royal blue) and a bag of 60 degree optics. I do actually have some 40 and 80 degree optics to experiment with but I think 60 will be the way to go.

http://hotimg23.fotki.com/a/70_70/254_23/LEDParts2-vi.jpg

I forgot to throw all the nice stainless steel bolts I picked up into the pictures. What I still need are fans when I decide exactly what the configuration of those will be and some Arduino bits to make a controller. The plan is to program it to gradually bring the blues from off to full illumination over 15 to 20 minutes and then gradually bring the whites up over another 15 to 20 minutes for a nice gradual sunrise. Sunset will be the opposite.

So now to start drilling and tapping the aluminum angle...

[StirCrazy]

what are you using for a soldering gun Ron?

Steve

[hillbillyreefer]

Should be a nice setup when your done. Could you keep us updated on costs and suppliers too.

[banditpowdercoat]

OOHHHH Nice :D

[bvlester]

very nice where did you order the kit from?

Bill

[Ron99]

Steve, just a plain jane handheld soldering iron. If I can remember where I put it when we moved in December :)

Hillbillyreefer and bvlester, the supplies came from different places. LEDs, drivers and optics came via the nano-reef.com group buy (better pricing that way). The housing was a freebie that was going into the trash elsewhere. Heatsinks were salvage/surplus from eBay (they aren't the prettiest but they were a third the price of buying a new heatsink and they'll do the job). Angle aluminum from a metal supply shop (Metal Mart; half the price for the same stuff at home depot etc. and they cut it to length for me). Stainless hardware came from Airarms (fastener supplier). The stainless bolts were like one third the price of the same thing at home depot :(

I really recommend buying supplies anywhere but one of the big box stores if you can. The stainless bolts were a great example. They were something like 40 cents each from an actual fastener supplier and $2.50 for a pack of 4 at Home Depot. Same for the aluminum angle. IIRC an 8 foot length at HD was $40 plus tax. Paid $20 at Metal Mart.

So all in so far I think I am in for around $1000 in parts, most of that being spent on the LEDs, optics and drivers. A couple of fans will cme from computer suppliers; probably NCIX or similar and the Arduino parts will probably be ordered from robotshop.ca.

[BlueAbyss]

Hmm...

LEDs, drivers, optics:

www.ledsupply.com - US
www.cutter.com.au - Australia, more choice

Heat sinks:

www.heatsinkusa.com - US

Arduino, mechanical bits, etc:

www.robotshop.ca - Canada

I'll be building an LED rig this summer, using Cree MC-E whites and XR-E blues. I have yet to decide which driver will be the best for my needs though, so I'm still in the research phase.

I'm actually considering building a plant growth rig with XP-G whites, partially to see how these new LEDs are and partially because I don't have enough window space for all my seedlings... maybe I'll build that first and put it on my 10 gallon when the seedlings move outside.

Either way, hope it goes well for you... can't wait to see how it works when you're done.

[Ron99]

Hmm...

LEDs, drivers, optics:

www.ledsupply.com - US
www.cutter.com.au - Australia, more choice

Heat sinks:

www.heatsinkusa.com - US

Arduino, mechanical bits, etc:

www.robotshop.ca - Canada

I'll be building an LED rig this summer, using Cree MC-E whites and XR-E blues. I have yet to decide which driver will be the best for my needs though, so I'm still in the research phase.

I'm actually considering building a plant growth rig with XP-G whites, partially to see how these new LEDs are and partially because I don't have enough window space for all my seedlings... maybe I'll build that first and put it on my 10 gallon when the seedlings move outside.

Either way, hope it goes well for you... can't wait to see how it works when you're done.

I would add nanotuners.com for LEDs and drivers as they have slightly better prices than ledsupply and also carry the Meanwell drivers which are a good choice. The Meanwells are nice because they come in dimming versions, run off AC power so there is no need for a separate power supply and can drive a good number of emitters.

Heatsink USA is not bad if you want a brand new heatsink. But the heatsinks I bought are actually wider at 12" than heatsink USA's widest at just over 10" and for a similar length of the 10" wide from heatsink USA it would have cost well over $200 with shipping. I have 45" x 12" of surplus heatsink at a total cost with shipping of $75. So keep your eyes on eBay as there are lots of used heatsinks available. If you are patient you can find what you need at a good price.

I'm not sure the MC-Es offer a huge advantage over the XR-Es. Yes, they have more lumens but they also generate more heat since they are essentially 4 emitters on one die. Also, are there any good choices in optics for MC-Es? I feel that the XR-Es with good optics provide more than enough PAR and if you need more for say a deeper tank than the XP-G may be a better option at this point.

[Ron99]

Some progress. Here are the two aluminum angles mounted in the fixture:

http://hotimg25.fotki.com/a/70_70/254_23/AlAnglePlacing-vi.jpg

And then I lay the three heatsinks on top to figure out spacing and where to drill my mounting holes:

http://hotimg25.fotki.com/a/70_70/254_23/HeatsinkSpacing-vi.jpg

Now because of the shape of the housing the aluminum angle had to be mounted pretty much flush with the bottom of the housing but I need clearance for the LEDs and optics so the heatsinks have to be raised a bit. So I cut some nylon spacers to place between the heatsinks and the angle.

http://hotimg23.fotki.com/a/70_70/254_23/Spacers-vi.jpg

And here it is all assembled:

http://hotimg23.fotki.com/a/70_70/254_23/MechAssemb-vi.jpg

So now that the mechanical engineering part is more or less sorted its on to the electrical engineering :biggrin:

[StirCrazy]

what are the milled squars for? is that the bottom or the top?

Steve

[Ron99]

what are the milled squars for? is that the bottom or the top?

Steve

The heatsinks came like that. Probably held some components or something. They are only 1/16 of an inch deep so a few emitters will sit 1/16 inch higher than the rest. I figured 1/16 inch won't make an appreciable difference in PAR and coverage etc. and given the cost savings of the heatsinks vs. something pristine I could live with it. It won't be visible inside the housing anyhow.

[Ron99]

Heatsinks cleaned and polished a bit and 80 emitters mounted:

http://hotimg23.fotki.com/a/70_70/254_23/LedsMounted-vi.jpg

next up, wiring...

[pangp]

tagging along..

looking good so far!

[BlueAbyss]

I would add nanotuners.com for LEDs and drivers as they have slightly better prices than ledsupply and also carry the Meanwell drivers which are a good choice. The Meanwells are nice because they come in dimming versions, run off AC power so there is no need for a separate power supply and can drive a good number of emitters.


Aha, forgot about nanotuners... they also carry some other cool stuff. LEDsupply carries Meanwell drivers also, I'm sort of leaning towards non-dimming since I no longer have the patience to experiment with PWM controls. That may change in the near future as I need a temperature controller and Arduino may be the ticket.

Heatsink USA is not bad if you want a brand new heatsink. But the heatsinks I bought are actually wider at 12" than heatsink USA's widest at just over 10" and for a similar length of the 10" wide from heatsink USA it would have cost well over $200 with shipping. I have 45" x 12" of surplus heatsink at a total cost with shipping of $75. So keep your eyes on eBay as there are lots of used heatsinks available. If you are patient you can find what you need at a good price.

I'm not sure the MC-Es offer a huge advantage over the XR-Es. Yes, they have more lumens but they also generate more heat since they are essentially 4 emitters on one die. Also, are there any good choices in optics for MC-Es? I feel that the XR-Es with good optics provide more than enough PAR and if you need more for say a deeper tank than the XP-G may be a better option at this point.

I really just meant to list some websites that I had found for such things. Ebay can definitely get you a better deal on just about anything (though I don't trust cheaply built Asian anything). I like the idea that people could salvage good aluminum heatsinks from computers and put them to good reuse as lighting equipment also (though this only works in nano-sized situations).

The MC-E's only advantage is that they are basically 4 XR-Es on a single die that can't be run as hard as the XR-E because of their higher 'thermal density'. There are just as many optic options for the MC-E as the XR-E. I am choosing them because they offer exactly that, more power with less space. If you're using more than 12 LEDs though, I doubt there is any reason to even glance at the MC-E. XP-Gs seem to be the most efficient, but any of the high powered LEDs will do the trick with the right optics (and optics are made for all of the popular LEDs), no matter the depth of the tank. I'm pretty sure that with 8 degree optics you could get good coverage on the bottom of a very deep tank.

My setup personally will consist of about 10 LEDs, so MC-Es = more light, less space when working with PCB stars.

EDIT: I meant to post this earlier this week but forgot... check out the 'Relative Spectral Curve's for the MC-E RGBW chips. http://ledsupply.com/docs/cree-mce1.pdf Shouldn't the curve for the 6.5K be heavier in blue than in red than 4K, rather than the other way around?

[StirCrazy]

I don't think the MCE are a good choice at this point. there white will realy over power blue and you won't be able to get as good of color variation as you can with the other ones right now.. well at least till they make a blue chip not a color chip. the color curves are right, remember a K temp can be made a thousand different ways so yes you could have a overall higher K temp with less blue as well as less red and green.

Steve

[BlueAbyss]

I only mentioned the RGBW in passing, I have no intentions of using them. I'll be using a 6.5K version. If they could fit 4 XR-Es into a 2" square, they would be a better option for my needs at the moment than the MC-E. I'm also considering using a 'light engine' and modifying it to my needs, rather than using the MC-E, but part of the reason I want to use the MC-E (with optics) is because it's something I haven't seen it done yet and I need this light source to be as compact as possible.

[Ron99]

Calvin, what are the dimensions o f the tank you are trying to light? The XR-E is plenty up to 24" with optics. 40 degree optics will probably put you between 250 watt and 400 watt metal halide PAR. I'm actually going to do some PAR testing when I have the array up and running so that will give us all some real numbers to look at.

As for heatsinks, eBay can be good for all sorts. Mine were definitely not from any computer. I am using three that are 12 inches by 15 inches and weigh 15 to 20 pounds each. they were industrial heatsinks of some sort. Lots of industrial heatsinks show up on eBay which is what I was looking for. You can find large heatsinks for any application on eBay, especially if you don't care how pretty they are. I was more concerned with size and functionality as they won't be visible in the housing. If they will be exposed and you want them to look nicer then Heatsink USA is a better, albeit more costly, option.

[BlueAbyss]

Yours are definitely not from a computer :lol: Curiously large heatsinks anyhow.

Hmm, I'll give you some background on my plan. I had a mysterious tank crash or illness take a few of my zoa colonies October / November, then I moved in December and lost my firefish and hermit crab. My goby is MIA, so assumed dead... So I'm left with 5 surviving coral colonies that I want to move into a second 10 gallon where I can better control the conditions (what I have now is pretty good for macros and crabs, but a little unstable for good coral growth).

The plan is to use a single 'spotlight' sort of light highlighting a rock island in the middle of the tank. Basically, an MC-E star surrounded by XR-E royal blue stars. I don't really think I'll use optics with them. I'll get to that with the 29 gallon tank later this summer, where I plan to use XR-Es and T5. I want to get away from halides, they project a lot of heat and I find it hard to deal with even with only a 70W.

EDIT: Found and read this tonight... very interesting. http://www.advancedaquarist.com/2010/2/aafeature2

[StirCrazy]

. Basically, an MC-E star surrounded by XR-E royal blue stars. I don't really think I'll use optics with them.
EDIT: Found and read this tonight... very interesting. http://www.advancedaquarist.com/2010/2/aafeature2

Blue, from everything I have read the MC-E washes out the XR-E RBs which is why no one is using the MC-E in builds. maybe build it so it is easy to change out if you get that happening and want more blue. would be interesting to see how the PAR is on a MC-E with 40 degree optics.:mrgreen:

I saw that article befor.. I don't believe either of the LEDs use optics so it is a good test for showing LEDs with no optices are just as good as T5s

Steve

[Ron99]

I would be really hesitant to use an MC-E over a 10 gallon tank. Even without optics you might burn your corals. Without optics you won't get much of a spotlight effect as the LEDs emit over about 120 degree angle. But even so they produce a hotspot under the emitter, even without optics. If you want a spotlight effect you would be better off using a small array of XR-Es with 60 degree optics. And as Steve said, the MC-E may wash out your blues. There is a reason that both DIY guys and commercial companies don't use the MC-E. I see what you're thinking but I have my doubts it will work as you plan.

[BlueAbyss]

I see what you're all saying. I'm sort of tempted to use a single MC-E as a plant grow light though, I'll bet it will make for some stocky seedlings (like metal halide)... something to explore. But I see what you're saying about having it over a 10 gallon... not enough depth, I suppose you would need a fairly deep tank to make it worth it.

Alright, XR-Es it is! I still want to run them at a lower mA though.

[Ron99]

Have the first row of LEDs wired up:

http://hotimg25.fotki.com/a/70_70/254_23/Row11-vi.jpg

Here's a close up shot:

http://hotimg23.fotki.com/a/70_70/254_23/Row1Close-vi.jpg

And the moment of truth...

The whites fired up:

http://hotimg25.fotki.com/a/70_70/254_23/WhiteTest-vi.jpg

And blues fired up:

http://hotimg25.fotki.com/a/70_70/254_23/BlueTest-vi.jpg

They're bloody bright! Can't look directly at them.

[The Codfather]

Nice.

[Ron99]

I see what you're all saying. I'm sort of tempted to use a single MC-E as a plant grow light though, I'll bet it will make for some stocky seedlings (like metal halide)... something to explore. But I see what you're saying about having it over a 10 gallon... not enough depth, I suppose you would need a fairly deep tank to make it worth it.

Alright, XR-Es it is! I still want to run them at a lower mA though.

I run XR-Es at 350mA on my 3 gallon pico. So you could run probably 500mA with optics and have enough PAR for SPS in a 10 gallon. I will be borrowing a PAR meter shortly and will measure PAR in the pico running 350mA and no optics for reference.

[StirCrazy]

hey Ron, have you decided how your going to run the maxwell dimming from the andurino? Never mind I see you used the PWM models. but how do you configure the signal from dimming in these ones?


Steve

[Ron99]

Still needs a bit of extra circuitry as the Arduino puts out a 5V PWM signal and the Meanwell needs 10V. So it requires a 12V DC supply, a voltage regulator to get 10V, and a transistor that applies the 5V PWM from the Arduino to the 10V line.

Now I imagine you will ask why not just use a 10V DC supply without the voltage regulator? Most 10V AC/DC converters are not 10V. Some can be less but often they can be more; closer to 12V even. But the Meanwell dimming circuit is very sensitive and anything above something like 10.6 or 10.8 V will fry it so I want to make sure I have a steady 10V for the dimming. I may also use the same 12V converter to supply 5V with a voltage regulator to power the Arduino too rather than have a separate 5V supply for that.

[donlite]

Hi, I own a company Sensa-Light Ltd. We do custom lighting. The LED set up you have looks good, but it looks like you have good heat dissipation from the heat sinks. If you can hook up a "K" probe ( most cheap ohm meters will have or use one) and test the heat you may be able to forget the noisy fans.

I am using 1 watt lumileds in blue and white combination to get about 14,000k
I think I have about 30 LEDS on a 55 gallon tank. The home made unit does not need any fans. Good luck, if I can help let me know, Don:idea:

[superduperwesman]

NICE! Want to build me a 6' one?? ahah

[rstar]

Tagging along! Keep up the good work, lookin good!

[Edmonton newbie]

looks great cant wait to see how that works for you, thinking i need something like that for my next tank

[Ron99]

Hi, I own a company Sensa-Light Ltd. We do custom lighting. The LED set up you have looks good, but it looks like you have good heat dissipation from the heat sinks. If you can hook up a "K" probe ( most cheap ohm meters will have or use one) and test the heat you may be able to forget the noisy fans.

I am using 1 watt lumileds in blue and white combination to get about 14,000k
I think I have about 30 LEDS on a 55 gallon tank. The home made unit does not need any fans. Good luck, if I can help let me know, Don:idea:

Thanks. The heatsinks are probably overkill but that's good. I will probably still run fans as they will be in an enclosed housing and I also want to put the drivers in the housing so it is all self contained. So extra cooling will be good. I plan to run the array for up to a few hours after it's all wired up and see how hot the heatsink gets.

NICE! Want to build me a 6' one?? ahah

Well, if you aren't in a hurry stay tuned as I have something brewing :wink:

Tagging along! Keep up the good work, lookin good!

Thanks. I can't wait to see how it turns out myself :lol:

[hillegom]

Ron, computer power supplys put out 5V and 12

[Ron99]

Ron, computer power supplys put out 5V and 12

It's a good idea but it's to big and overkill for the need. I will just need a small 12v wall wart and then the voltage rectifiers. That way it will fit in the housing. I wouldn't be able to squeeze a computer PS into it. I can actually probably run my fans off it too as the Arduino and PWM circuitry won't draw alot of current.

[hillegom]

How many amps at 12V do you need?

[StirCrazy]

Still needs a bit of extra circuitry as the Arduino puts out a 5V PWM signal and the Meanwell needs 10V. So it requires a 12V DC supply, a voltage regulator to get 10V, and a transistor that applies the 5V PWM from the Arduino to the 10V line.



um, thats stupid, I figured the PWM models would be based on 5V as there is only need for a pulsed signal not a voltage.. oh well.

Steve

[StirCrazy]

Ron, what about using a 10V power supply to power the arduino, and then taking the power from your 9V out wich will actualy be 10V. this will eliminate that need of another power supply as I can't imagin you need much of any current for the PWM.

Steve

[Ron99]

How many amps at 12V do you need?

Only a couple of amps. The Arduino itself uses very little current, it dependes on what is plugged into it. I will probably only be using it to generate the PWM signal and power a real time clock circuit so it will likely only be using a few hundred milliamps at most. the three fans will probably use more and fans typically draw around 300mA; maybe a little more. So a 2A power supply will likely be more than enough.

um, thats stupid, I figured the PWM models would be based on 5V as there is only need for a pulsed signal not a voltage.. oh well.

Steve

Yeah, but I guess they did not design it with microcontrollers in mind. You can apparently use lower voltage but it will not go to full brightness. I am using a 9V battery for test purposes just to get it to fire up. It lights but is probably not a full brightness. It is basically using a 10V signal for full bright that is then chopped up for the dimming.

Ron, what about using a 10V power supply to power the arduino, and then taking the power from your 9V out wich will actualy be 10V. this will eliminate that need of another power supply as I can't imagin you need much of any current for the PWM.

Steve

You are correct. I am still figuring out the Arduino and assumed it needed 5V in but it can take 12V in as it has it's own voltage regulator on board. So I can use a single 12V supply for the fans, the Arduino and then with a voltage regulator to supply the PWM. I don't see a 9V out on the Arduino, just a 3.3V and 5V outputs.

[StirCrazy]

You are correct. I am still figuring out the Arduino and assumed it needed 5V in but it can take 12V in as it has it's own voltage regulator on board. So I can use a single 12V supply for the fans, the Arduino and then with a voltage regulator to supply the PWM. I don't see a 9V out on the Arduino, just a 3.3V and 5V outputs.

there is a pin labled Vin it can be used for a hardpoint voltage supply or you can take off there to power something else and it will be what ever power you are feeding your board with.. so if you are using a 12v adapter (which is probably what you should use. it will be 12V out. then you can step it down to 10V as you lose about 0.7 of a volt. so that out with a 12V adapter will actualy give you about 11.3V

Steve

[Ron99]

there is a pin labled Vin it can be used for a hardpoint voltage supply or you can take off there to power something else and it will be what ever power you are feeding your board with.. so if you are using a 12v adapter (which is probably what you should use. it will be 12V out. then you can step it down to 10V as you lose about 0.7 of a volt. so that out with a 12V adapter will actualy give you about 11.3V

Steve

I guess it's 6 of one half dozen of another. I still need to run that through a voltage regulator and then a transistor switched by the 5V PWM from the Arduino. So whether it originally comes off the 12V out from the Arduino or right off the 12V power supply probably makes little difference functionally. Although I may be able to do something more compact with a proto shield on the Arduino itself using the 12V Vin as you suggest.

[Ron99]

Quick update. Got all the LEDs wired and test fired. All work fine. Now I need to get some connectors and wire up all the drivers. I am also short a few 60 degree optics. For some reason I thought I had more of them so I have to order up a few more. Arduino parts are ordered and I should have them by the end of the month.

[StirCrazy]

hey Ron, where are you ordering your Ardunio parts from?

Steve

[Ron99]

robotshop.ca which has a 5% off coupon code for the next few days.

[Newbies]

Just tagging along for future reference.... cool project
________
ford edis picture (http://www.ford-wiki.com/wiki/Ford_EDIS)

[donlite]

I will take some photos of my led array and post them.:idea:

[hillegom]

Ron, I have a 12V 3A power supply. 4.5 x 2.25 x 1.25 inches approx.
I would trade for a sps frag or whathaveyou

[Ron99]

Ron, I have a 12V 3A power supply. 4.5 x 2.25 x 1.25 inches approx.
I would trade for a sps frag or whathaveyou

Thanks Dietmar. I have a box of old AC/DC adapters so if I don't have a suitable one in there I'll definitely take you up on your offer. I have a few frags I can spare :smile:

[hillegom]

Sounds good. Except, I wouldn't want the frag right away. I still have that 120 to build. When it gets done, things would even out. You get the PS now, I get the frag later.
Its only taken more than a year to build the stand!

[Ron99]

After lots and lots and lots of wiring I have all the LEDs working. I decided to get fancy and do everything with molex connectors so stuff can be removed/replaced easily. But it looks a bit like wiring spaghetti right now. I'll clean it up as I mount it all in the housing. Here's the drivers and wiring and all the LEDS wired up:

http://hotimg23.fotki.com/a/70_70/254_23/Wiring-vi.jpg

http://hotimg23.fotki.com/a/70_70/254_23/Drivers-vi.jpg

http://hotimg23.fotki.com/a/70_70/254_23/AllWired-vi.jpg

And here it all is fired up.

http://hotimg23.fotki.com/a/70_70/254_23/AllOn-vi.jpg

Still need to mount optics, pick up some fans and I'm waiting for the Arduino parts which should arrive this week and I can figure out the controller.

[superduperwesman]

AWESOME! Looks very nice

[Zoaelite]

Ron that's pretty impressive! Watch out you might get sued for copy right infringement considering what happened to solaris ;).

I can't want to see it on an actual fish tank with with some PAR values!
Levi

[Ron99]

ron that's pretty impressive! Watch out you might get sued for copy right infringement considering what happened to solaris ;).

I can't want to see it on an actual fish tank with with some par values!
Levi
t

[Ron99]

Whoops, not sure what happened with that last post :redface:

AWESOME! Looks very nice

Ron that's pretty impressive! Watch out you might get sued for copy right infringement considering what happened to solaris ;).

I can't want to see it on an actual fish tank with with some PAR values!
Levi

Thanks guys. I say bring it! I'll take on Orbitech and their marginal patent!

So just out of curiosity, in air, 18 inches away from the LEDs and no optics yet: 260 PAR. This is gonna be good :mrgreen:

[Dyspnea]

Tagging along!

Very interesting thread :)

[Ron99]

For some added LED info, I just did some PAR readings in the iReef pico. it has 6 Cree XR-E LEDs being underdriven by a 350mA dimming buckpuck that I actually have turned down a little from full power. No optics are used, just bare emitters. PAR readings were about 170 near the top of my rock which is about 4 inches below the surface of the water and about 5 inches below the LEDs. Par drops to about 80 near the bottom of the tank about 10 inches below the water line.

[bvlester]

just wait till you can crank it up and let it go it should be nice and bright under there.

Bill

[Ron99]

A bit slow as I need to redo a few of the molex connectors and pick up a few little bits for assembly. But I did a quick test in air again with optics. At 18" with 60 degree optics the PAR jumps to about 375 vs 250 without optics.

[byee]

Still needs a bit of extra circuitry as the Arduino puts out a 5V PWM signal and the Meanwell needs 10V. So it requires a 12V DC supply, a voltage regulator to get 10V, and a transistor that applies the 5V PWM from the Arduino to the 10V line.

Now I imagine you will ask why not just use a 10V DC supply without the voltage regulator? Most 10V AC/DC converters are not 10V. Some can be less but often they can be more; closer to 12V even. But the Meanwell dimming circuit is very sensitive and anything above something like 10.6 or 10.8 V will fry it so I want to make sure I have a steady 10V for the dimming. I may also use the same 12V converter to supply 5V with a voltage regulator to power the Arduino too rather than have a separate 5V supply for that.

Ron99,

The max voltage for the Arduino is 9VDC. There is also an onboard voltage regulator which will regulate the voltage down to 5VDC for the electronics.

May need the use of OP amps to drive the Meanwell LED driver. Vancouver Reefer (Canreef member) may have something since he's been down this path with his setup. I believe he used the LED Pucks which is similar to the Meanwell.

There are also other threads on about LED lighting on some of the other aquarium forums.

I've also been toying with the idea of using the Arduino as a platform for my aquarium controller - pH, temperature and lighting.

Good luck!

[Ron99]

The Arduino specs say 6 - 20 volts as the limits for input with 7 - 12 volts recommended. I have a 12V power supply that will feed the Arduino, three 12V fans and, with an LM317 voltage regulator, the 10V for the PWM signal to the Meanwells (via a transistor controlled by the PWM out of the Arduino) so that is all figured out. Just have to build that part :). The main thing I need to figure out will be programming the Arduino as I'm not a programmer.

I have followed many threads on LEDs on reef central and nano-reef and have built a basic array for my pico tank already (6 LEDs and one dimming buckpuck). The array itself is all wired and works fine without dimming right now. Now it's on to assembly in the housing and then building and programming the controller.

[Ron99]

P.S. Big thanks to hillegom for the power supply. I owe you a frag or three when your new tank is up and running:thumb:

[hillegom]

Thanks Ron.
Will look forward to a frag or so. Nice to be able to help with that big project of yours.
Surplus here for surplus there.

[byee]

Also check out reefprojects.com for code.

Pointless to re-invent the wheel especially when all this stuff is open source.

[StirCrazy]

Ron99,


.

Good luck!

never mind Ron said it.

Thanks for the link though.

Steve

[Ron99]

Thanks for the link Byee, I'll check it out and see if they have useful code for me. I have an LCD input shield for the Arduino which has a small LCD display and several buttons so I would like to make it a self contained adjustable controller rather then just uploading code to it to make changes to settings.

[Skimmerking]

Ron where from NFLD are you from

[Ron99]

Ron where from NFLD are you from

The far west coast of NFDL, White Rock BC :lol:

I was actually born in Montreal, grew up in Toronto and have been in BC for over 18 years now.

[OceanicCorals-Eugene-]

Ron let me know if theres any help i can give you regarding the arduino coding, i have an army of coders at my disposal, not... but let me know if theres any part of the code you cant seem to get and maybe i can send some help your way.

[Ron99]

Thanks Eugene, I appreciate the offer. I am just getting to the Arduino part this week so will see how it goes.

[byee]

Thanks for the link Byee, I'll check it out and see if they have useful code for me. I have an LCD input shield for the Arduino which has a small LCD display and several buttons so I would like to make it a self contained adjustable controller rather then just uploading code to it to make changes to settings.

Ron99,

No worries! Glad I was able to help.

The whole reason I bought my Arduino was because of reefprojects.

I've had mine for about 9 months now but just 2 busy with stuff for around the house.

[Ron99]

So after struggling with the controller for a bit I realized that it is going to take me a little while to sort that part out. I have also been getting some algae and cyano showing up on my rocks which I am fairly certain is due to my old T5 bulbs (15+ months old) I decided to just set it up without the controller for now and I can add that in later.

Here's a pic of the enclosure wired up with fans and drivers:

http://hotimg23.fotki.com/a/70_70/254_23/EnclosureWired-vi.jpg

I also took a picture of the tank with the T5s using manual settings on the camera. Other then cropping a bit I didn't enhance the pictures at all as I wanted it to be a comparison between the two lights:

http://hotimg23.fotki.com/a/70_70/254_23/TankwT5s-vi.jpg

And here it is with the LEDs using the exact same camera settings:

http://hotimg23.fotki.com/a/70_70/254_23/TankwLEDs-vi.jpg

It's definitely more blue with the LEDs and to me the coral colours look a bit nicer. It is also a bit dimmer to the eye so lumens are probably less but that is normal with blue LEDs. However, here's the important info...PAR readings.

With the 8 x 54W T5HO fixture I had about 300 PAR near the top of the tank, around 240 or so near the middle and about 160 at the bottom. This is running the lights about 3.5 to 4 inches above the water.

Now with the LEDs at about 7.5 to 8 inches above the water PAR was a bit over 500 near the top of the tank, about 400 in the middle and about 290 at the bottom. Holy photons Batman!

So I raised the fixture to about 15 inches in height for now as I can't dim the LEDs yet. Par is now about 350 at the top of the tank, 280 in the middle and still around 200 at the bottom of the tank. Remember, this is with the fixture 15 inches above the tank:

http://hotimg23.fotki.com/a/70_70/254_23/LEDsHigh-vi.jpg

Also, the shimmer is really noticeable compared to the flatter T5 lighting. Oh and the bloody fixture weighs at least 50 pounds if not a bit more :surprise: I think I will pick up some heftier S hooks to match the heftier chain I have already switched to!

[freezetyle]

Great work! looks really promising

[burrows14]

WOW! Looks great Ron. Im very impressed:biggrin:

[StirCrazy]

Now with the LEDs at about 7.5 to 8 inches above the water PAR was a bit over 500 near the top of the tank, about 400 in the middle and about 290 at the bottom. Holy photons Batman!

So I raised the fixture to about 15 inches in height for now as I can't dim the LEDs yet. Par is now about 350 at the top of the tank, 280 in the middle and still around 200 at the bottom of the tank. Remember, this is with the fixture 15 inches above the tank:



which optics did you go with on the LEDs?

Steve

[Ron99]

Thanks guys. I'm happy to have it running as the T5s were overdue for replacement but I didn't want to spend over $250 on new bulbs when I was working in this.

Steve, I used 60 degree optics. Actually, a few emitters at the front of the array have 80s and the rest of the front row do not have any optics yet. I was short a number of 60 degree ones which I have on order but haven't received yet.

[derekc85]

nice build:

if you dont mind, can you share the parts list for LED's, drivers, config etc?

how much did it cost you at the end? are the LED's replaceable?

[Ron99]

Most of the details are in the thread but I'll summarize again:

80 x Cree XR-E LEDs; 40 cool white Q5 bin and 40 royal blue.
8 x Meanwell ELN60-48P drivers. Each is capable of driving up to 13 LEDs. 4 are driving the blues and 4 for the whites.
60 degree optics for the Cree LEDs
3 x 12"x15" eBay surplus/salvage heatsinks
1 Free surplus Hamilton Lighting housing
Lots of misc. wire, connectors, screws, aluminum etc.
An Arduino microcontroller and associated electronic bits to make the controller - still to be completed.

My costs are somewhere between $1200 and $1400 but I haven't done an exact accounting.

Since the drivers are capable of handling an additional 24 LEDS I will probably add a few 3 or 4 near UV (395nm) LEDs as an experiment and to bring out the coral florescence a bit. I will most likely run them without optics as I want them more for some supplementary wavelength rather than producing PAR. I also read a bit about using some green LEDs to bring out coral colours and am toying with the idea of similarly adding some green ones without optics for that purpose.

[StirCrazy]

Since the drivers are capable of handling an additional 24 LEDS I will probably add a few 3 or 4 near UV (395nm) LEDs as an experiment and to bring out the coral florescence a bit. I will most likely run them without optics as I want them more for some supplementary wavelength rather than producing PAR. I also read a bit about using some green LEDs to bring out coral colours and am toying with the idea of similarly adding some green ones without optics for that purpose.

Have you thought of going full UV instead of near? I think the UV is something were missing in the Led light set ups and they will make your corals pop like crazy. as for the green, I wouldn't waist my time, unless you got nothing better to do :mrgreen: I think green will have the same results of when people were trying the red ones mixed in. lead to more algae.

Steve

[Skimmerking]

So you are saying there Ron that those lights have better PArs then the 8x 54 w t-5's and what is the life span of the LEd's and if one blows how easy is it to change.

really good work man.

[Ron99]

Have you thought of going full UV instead of near? I think the UV is something were missing in the Led light set ups and they will make your corals pop like crazy. as for the green, I wouldn't waist my time, unless you got nothing better to do :mrgreen: I think green will have the same results of when people were trying the red ones mixed in. lead to more algae.

Steve

Thought about it Steve but the near UV ones at 395nm were had for $15 vs. closer to $100 for true UV-A at 365nm. I'm looking at options for 365nm but until the cost comes down I won't be trying them. As for the green ones, it's just a notion right now. I'll probably let the nano-reef.com crowd experiment with that first :biggrin: However, I don't think a bit of green will have the same effect on Algae that red would. The red is really into the spectrum algae likes for photosynthesis. Green is much closer to the blue end of the spectrum. If I did it I would only use maybe six emitters with no optics just too add some colour to the tank and help the red and orange pigments pop a bit. the blue and UV are great for the green pigments and to some extent the red but green apparently really brings out the reds and oranges.

So you are saying there Ron that those lights have better PArs then the 8x 54 w t-5's and what is the life span of the LEd's and if one blows how easy is it to change.

really good work man.

Yes, better PAR than my fixture which is a Sunlight Supply Sunblaze fixture. It's a pretty good one with contoured reflectors etc. but I am sure some of the others with true individual reflectors and active cooling will give a bit more PAR than mine. It was doing a great job and SPS were growing just fine under it. However, the LEDs have really blown it away with nearly twice as much PAR with about half the energy consumption. That's the beauty of LEDs as a light source. They are way more efficient than other methods.

Their life span, as long as they are adequately cooled, should be 50,000 hours. At 8 hours per day of lighting that is nearly 8 years. I am saving $250x8= $2000 in bulb changes alone not to mention using less electricity. Changing them is not that hard. Unsolder the wires and pry it off the heatsink (I used adhesive pads but if I had drilled and tapped the heatsink then it would just be a matter of undoing the screws holding the LED down). New adhesive pad and new LED and resolder the wires and you're good to go.

[Canadian]

Great build. And you've done a great job of documenting things for everyone to mimic should they so choose.

But I have to comment on the PAR. Your old T5 fixture is not a very good comparator for a moderately good T5 fixture. Your PAR values with the T5 fixture are abysmal and not reflective of typical high quality T5 fixtures with 54W lamps (presumably your reported T5 PAR values are with new lamps as well given you're comparing this to brand new LEDs). Secondly, the PAR values with your DIY LED fixture are still quite low. Every time a discussion about LEDs comes up the comparison to 250W and 400W MH is made. These numbers clearly show they aren't even close in the manner in which you've designed this fixture. I've maintained that until real numbers for PAR are measured and reported the propaganda of "equal to a 400W MH" is absurd. So far the data I've seen do not support the marketing BS. In all honesty, did you expect the PAR values to be that low?

And as far as lifespan goes: again there is no long term data for PAR as far as I know (unless you have info I haven't seen yet). T5 lamps have a reasonable lifetime but their effective PAR drops much faster. At this point we don't have any good data to support the numbers being batted around about 50,000 hours lifespan. Sure, the emitter may continue to emit light for 50,000 hours but at what PAR?

[StirCrazy]

But I have to comment on the PAR. Your old T5 fixture is not a very good comparator for a moderately good T5 fixture. Your PAR values with the T5 fixture are abysmal and not reflective of typical high quality T5 fixtures with 54W lamps (presumably your reported T5 PAR values are with new lamps as well given you're comparing this to brand new LEDs). Secondly, the PAR values with your DIY LED fixture are still quite low. Every time a discussion about LEDs comes up the comparison to 250W and 400W MH is made. These numbers clearly show they aren't even close in the manner in which you've designed this fixture. I've maintained that until real numbers for PAR are measured and reported the propaganda of "equal to a 400W MH" is absurd. So far the data I've seen do not support the marketing BS. In all honesty, did you expect the PAR values to be that low?

And as far as lifespan goes: again there is no long term data for PAR as far as I know (unless you have info I haven't seen yet). T5 lamps have a reasonable lifetime but their effective PAR drops much faster. At this point we don't have any good data to support the numbers being batted around about 50,000 hours lifespan. Sure, the emitter may continue to emit light for 50,000 hours but at what PAR?

Andrew, you have to go do some reading in this thread, he is getting higher par at 5 times the distance. T5 was at 3.5" and he is still getting higher PAR at 17" above the tank, which he did becasue he didn't have his dimming setup done yet. At 8" he was over 500 PAR where at 3.5 with the T5 he was at 300 PAR, Also this is with 60 and 80 degree optics, no one ever said thoes optics will give you 250 - 400 MH levels, it is 40 degree optics you need to get that, the best you will get with 60's is around a 150 watt MH which he is pretty close to as a 150watt mh isn't much higher than a good T5 set up. if you want a good comparason maybe Ron will take a PAR reading at 3.5" with the LEDs then you can see how much more punch he is realy getting, I am going to guess that he will be around 700 at the surface and probably close to or over 400 at the bottom, but thats a guess.

now as for the PAR output, you are thinking along the lines of a gas filled cathode tube. you can't think that way with LED as there is no gas to break down and change the spectrum. it has been showen there is absolutly no shift in spectrum in a LED over its life, only a 15% decrease in intensity, so there is no reason to think other than a 15% decrees in PAR as there will be no spectrum shift to compound the drop as it does in MH, PC, CF, ect. so if we look at that the average MH has a 20% drop by the time it is changed, and most of that drop occures int he first 6 months, floressents have a even steaper drop off, so going with LEDs after 5000 days (10 hours per day) you will have 15% less intensity with no spectral shift so you should not get nusence algae ect..

Steve

[Ron99]

I think Steve pretty much summed it up. Just to add to it, my PAR reading on the T5 were with the old bulbs so yes, the reading would have been higher with newer bulbs. Couldn't do that as I just recently borrowed a PAR meter for this testing and had no way to test when my bulbs were new. And I wasn't about to spend $250+ on new bulbs at this point :) I would love to see others posting PAR numbers from their fixtures so we can do comparisons.

And for the record, I wasn't making any claims as to comparisons with 250w or 400W MH as I haven't done any PAR readings from those but a guess as to where my light output would end up. If anybody else has let's see the numbers as it would be interesting to compare.

But what I do know is that with the LEDs lower down, at about 7 to 8 inches above the water I was seeing just over 500 PAR in the upper part of the tank and 300 just above the sand bed of the tank about 18 inches down. That's pretty good from my understanding. I didn't test with the fixture even closer to the water but if I lowered it to where I had the T5s then I suspect Steve is correct and PAR would be 600 to 700 in the upper part of the tank. Maybe I'll try it tomorrow and see. The reason I raised the fixture for now is that I would burn all my corals switching them from say 300 PAR to 500+ PAR. They need to be acclimated and I plan to slowly lower the fixture every few days to acclimate them to the higher light.

Give this article a read:

http://www.advancedaquarist.com/2010/2/aafeature2

It shows comparisons in air, not water but it's a start. Also you have to note that the LED fixtures used in the article are not the best. One is a PFO Solaris which used older LEDs that do not have the same output as the newer generations of emitters and the other is the Eco-Lamps one that under drives their LEDS at just over 1 watt instead of 3 watts. Even so they compared well to a high end T5 (about equal performance) and a 250W MH (a little less but not by much; about 15 to 16% less output). I guarantee you my fixture will outperform both the Solaris and the Eco-Lamps as I used very good Cree emitters driven at a full 3W and 1000mA for the whites and about 900mA for the blues.

Even so Dr. Joshi states that most Acropora and light loving corals will thrive at PAR levels of 300 to 400 which this fixture can provide throughout my tank if I wish. That's good enough for me. In a 30" deep tank for example you could keep high light SPS in the top 2/3 of the tank. I don't have a deeper tank to test on but I suspect PAR would still be close to 200 by 30" down. That's pretty good performance in my book.

[Ron99]

For comparison purposes Eugene at Oceanic Corals tested their 400W MH fixture (in their LED spotlight thread) and says that PAR was 500 at 12" below the water. I had 400 PAR at about that depth with the fixture say 7.5 inches above the water. I'll try to test with the fixture closer to the water at 12" depth and see what I come up with.

But I think my 400PAR as tested so far shows pretty well against the MH.

[Ron99]

Oh, and just to Spam my own thread a bit testing by the LED leaders at nano-reef.com shows that the drop off in PAR vs. depth is less with LEDs vs. MH. That makes sense as with the optics virtually all the output of the LEDS is focussed down into the tank. T5s and MH, even with good reflectors, probably have less cohesive light with more scatter from the reflectors as they emit in all directions and then rely on the reflectors to focus the light down if that makes sense.

[Canadian]

Andrew, you have to go do some reading in this thread, he is getting higher par at 5 times the distance. T5 was at 3.5" and he is still getting higher PAR at 17" above the tank, which he did becasue he didn't have his dimming setup done yet. At 8" he was over 500 PAR where at 3.5 with the T5 he was at 300 PAR, Also this is with 60 and 80 degree optics, no one ever said thoes optics will give you 250 - 400 MH levels, it is 40 degree optics you need to get that, the best you will get with 60's is around a 150 watt MH which he is pretty close to as a 150watt mh isn't much higher than a good T5 set up. if you want a good comparason maybe Ron will take a PAR reading at 3.5" with the LEDs then you can see how much more punch he is realy getting, I am going to guess that he will be around 700 at the surface and probably close to or over 400 at the bottom, but thats a guess.

now as for the PAR output, you are thinking along the lines of a gas filled cathode tube. you can't think that way with LED as there is no gas to break down and change the spectrum. it has been showen there is absolutly no shift in spectrum in a LED over its life, only a 15% decrease in intensity, so there is no reason to think other than a 15% decrees in PAR as there will be no spectrum shift to compound the drop as it does in MH, PC, CF, ect. so if we look at that the average MH has a 20% drop by the time it is changed, and most of that drop occures int he first 6 months, floressents have a even steaper drop off, so going with LEDs after 5000 days (10 hours per day) you will have 15% less intensity with no spectral shift so you should not get nusence algae ect..

Steve

Steve,

You'll have to read Ron's post after yours. The PAR measured with the low quality T5 fixture was on more than 1 year old T5 lamps and therefore a very poor comparator - forget about the height of the fixture comparison. Additionally, I'm less concerned about the comparison to T5s than I am to MH (the only reason I commented on T5s was because the comparison was hugely favorable to the LEDs based on the manner in which it was performed). Various qualities of LEDs make them favorable to T5s IMO. With that said, comparisons are being made all over the place to MH by retailers, manufacturers, and hobbyists. When you were standing on your soap box about MH a few months ago you pounded your chest about getting over 1000 PAR at the surface and 500 on the bottom of the tank. Clearly, the numbers being reported here don't fall in line with what you've purported constitutes adequate PAR for your SPS tanks.

And as far as the theory behind intensity drop in LEDs goes - it's just that: a nice theoretical expectation based on presumed qualities. Until there is some long term data for PAR the theoretical musings can carry on but with a healthy dose of skepticism.

[Ron99]

Steve,

You'll have to read Ron's post after yours. The PAR measured with the low quality T5 fixture was on more than 1 year old T5 lamps and therefore a very poor comparator - forget about the height of the fixture comparison. Additionally, I'm less concerned about the comparison to T5s than I am to MH (the only reason I commented on T5s was because the comparison was hugely favorable to the LEDs based on the manner in which it was performed). Various qualities of LEDs make them favorable to T5s IMO. With that said, comparisons are being made all over the place to MH by retailers, manufacturers, and hobbyists. When you were standing on your soap box about MH a few months ago you pounded your chest about getting over 1000 PAR at the surface and 500 on the bottom of the tank. That's all I am trying to say. I would love to compile Clearly, the numbers being reported here don't fall in line with what you've purported constitutes adequate PAR for your SPS tanks.

And as far as the theory behind intensity drop in LEDs goes - it's just that: a nice theoretical expectation based on presumed qualities. Until there is some long term data for PAR the theoretical musings can carry on but with a healthy dose of skepticism.

First I think you are being overly critical of my light fixture. I would not call it low quality but about a middle of the road one. Sunlight Supply makes good fixtures and these were not much different then the first generation Tek fixtures in terms of lighting output. Slightly better reflectors and active cooling would improve it somewhat but how much do you think that would be? 10%? 20%? Secondly, while the bulbs were old, the only references I could fine say that T5s lose about 10% output over 2 years. The problem for reef tanks is the shift in colour which can lead to undesirable algae etc. But let's say that the degradation was 20% for the sake of argument. So my 300 PAR with the T5s could mean 375 PAR with new bulbs. Still much less than 500 PAR with the LEDs higher above the tank then the T5s were. Argue all you want but however you set up my test, new bulbs or not etc. my LEDs handily outperform my T5s. That's all I am claiming. I would love to test different lighting setups for comparison but I don't have easy access at the moment. Maybe we can do that and arrange to test various people's lights with the same meter which will give some idea. it's not as good as a side by side comparison under the same conditions in the same tank but it would be a start.

I would also take manufacturers claims with a grain of salt as they are in the business of marketing their products and most likely set up tests to skew results in favour of their products. However, independent test are being done like Sanjay Joshi's comparison I linked in my last post. That clearly showed the "lower quality" LED fixtures performing as well at or slightly better then a high end T5 fixture and only 15 to 16% lower output then a good 250W MH setup. Dr. Joshi is an engineer and has been testing all sorts of lights, especially MH, for a long time so I would consider him a good independent reviewer. Here's also a more recent review of a top of the line LED fixture, the AquaIllumination:

http://www.advancedaquarist.com/2007/10/review/view?searchterm=PAR (http://www.advancedaquarist.com/2007/10/review/)

This one clearly shoes the AI unit outperforming a 250W MH unit. Granted, the MH ballast used may not be top of the line but is probably typical of one used by many people.

With respect to who has the biggest PAR; really, who cares. Stoney corals appear to photosaturate between 400 and 500 PAR so having 700 or 1000 PAR is probably pointless other than for bragging rights. I doubt it will do the corals much good.

As for longevity and degradation of LEDS, that is fairly well established. While high powered LEDs are fairly new to our hobby they have been around for a decade and are well tested. Low power LEDs have been around far longer then that so the degradation of LEDs is well understood.

[Dyspnea]

You have done a great job, look amazing! Still debating myself whether i should buy myself at 4 foot MH fixture or be adventorous and build an LED setup myself.

[Ron99]

You have done a great job, look amazing! Still debating myself whether i should buy myself at 4 foot MH fixture or be adventorous and build an LED setup myself.

Thanks. I say go LEDs. You won't regret it. Lots of people are having great success with DIY LED setups.

And just to throw more fuel on the fire I should add that that last review I linked to was a test of the first generation AI fixture. The second generation should perform even better as it uses the latest Cree XP-G LEDs that have higher output than the SSC ones used in the first gen fixtures.

[Ron99]

Played around lowering the fixture and found that the colour blending and overlap of the emitters wasn't great in the upper portions of the tank with the lights 3 inches above the water. PAR was between 550 and 600 in the upper part but stayed at 500 in the middle of the tank and was close to 400 at the bottom. I think that to get optimal blending of the light from the emitters and good overlap 6 to 7 inches above the tank is about as low as I will want to run them.

[Dyspnea]

What are you planning for livestock? SPS, LPS, clams, zoa, softs... etc?

[StirCrazy]

Steve,

. When you were standing on your soap box about MH a few months ago you pounded your chest about getting over 1000 PAR at the surface and 500 on the bottom of the tank. Clearly, the numbers being reported here don't fall in line with what you've purported constitutes adequate PAR for your SPS tanks.
.

yup, I did, and I have no doubt that if Ron put 40 degree optics on his fixture, at 3" he would blow away my MH. and just to add to it my MH were unusaualy high output which was a combanation of the right ballast, my own design for the reflector and good bulbs. in the 25 MH setups I tested the only one that came close was a 400 watt sona agro ballast running iwasaki 6500K bulbs and it was still only slightly over 800.

so I can safly say that a good LED set up such as Ron's will blow any T5 out of the watter, most 150 watt setups, and some of the 250 watt setups. if he would have gone a little tighter spacing which allowed for 40 degree optics with out spotlighting I can say you would be hard to find a MH system that would be a higher output, and it would be pointless to try as at that output level would it realy matter?

Is this because you spent to much on your T5 set up and are haveing buyers regrets Andrew? :mrgreen: don't worry I spent close to 10K on lighting in 2 years befor I found something I was happy with myself. now I did recover some of that selling reflectors and reflector material, but still spent way to much on lighting.. thats why the new tank is going to have LED lighting, as it will give me the ability to change color K value on a whim, and also talyor the intensity to what I have in the tank at the moment. if I get a bunch of new frags, I can dial it down and have it increase slowly over a week or two to prevent bleaching. I can have a gradualy sunrise/sunset instead of a two stage on/off. plus the most fun is building the stuff now keeping the tank going.. maybe I should just build systems for people who want a tank but dont want to build it :mrgreen: then I could just keep on doing the fun stuff .

Steve

[Ron99]

What are you planning for livestock? SPS, LPS, clams, zoa, softs... etc?

All of the above:biggrin: I currently have a mixed bag in the tank with SPS, LPS, softies, and zoas. I hope to be able to keep all of them with careful positioning. But I already found the T5s were to much for some of the LPS and had to tuck them under overhangs etc. I would like to add a small clam if space permits but we'll see.

Also, this latest news as of today is interesting. LEDs are only going to get better while MH and T5s etc. are pretty much as good as they are going to get now:

Cree, Inc. (Nasdaq: CREE), a market leader in LED lighting, announces a breakthrough new lighting-class LED platform, the XLamp® XM LED. This new single chip LED delivers record-breaking efficacy of 160 lumens per watt at 350 mA. The LED also delivers 750 lumens at 2 A, which is equivalent to the light output of a 60 W incandescent light bulb at less than 7 watts.

“This new platform continues Cree’s well-established record of turning R&D innovations into products,” said Chuck Swoboda, Cree chairman and CEO. “We continue to set the pace for LED performance, establishing new benchmarks that make you wonder why anyone would consider last-century’s energy-wasting technology.”

A cool white XM LED driven at 350 mA can produce 160 lumens at 160 lumens per watt. The new platform has a larger footprint than Cree’s XP family and also offers the unique combination of very high efficacy at very high drive currents. At 2 A, an XM LED produces 750 lumens at 110 lumens per watt. The thermal resistance of the XM platform is 2 degrees C per watt— an industry-leading technology breakthrough and a 350 percent improvement over Cree’s flagship XLamp XP-E LED.

Samples of the XLamp XM LEDs are available for order with standard lead times and commercial availability is targeted for Fall 2010.

[blacknife]

I so need to build my own light when i decide what my next tank is.. it looks like a fun project. and when things go wrong.. there is only one person to blame.. yourself.

[Dyspnea]

Where did you find most of your reading and research?

[Canadian]

I apologize in advance for the odd timing of these posts. I had intended to post after work but got caught up with the realtor and was running around all night until going to bed. Now I'm awake in the middle of the night so I thought I'd do something "productive" :)

First I think you are being overly critical of my light fixture. I would not call it low quality but about a middle of the road one. Sunlight Supply makes good fixtures and these were not much different then the first generation Tek fixtures in terms of lighting output. Slightly better reflectors and active cooling would improve it somewhat but how much do you think that would be? 10%? 20%?

Nope. It's low end. For comparison here is a picture with PAR values for a higher quality T5 fixture. Keep in mind the setup here is still far from ideal given that the fixture doesn't extend the length of tank so keep your eye on the PAR values in the middle of the tank. In this picture you're seeing similar PAR values to your LED fixture with newer lamps and better quality fixture.

http://i43.photobucket.com/albums/e355/kangym/FTS7-3-09-1.jpg

Secondly, while the bulbs were old, the only references I could fine say that T5s lose about 10% output over 2 years. The problem for reef tanks is the shift in colour which can lead to undesirable algae etc. But let's say that the degradation was 20% for the sake of argument. So my 300 PAR with the T5s could mean 375 PAR with new bulbs. Still much less than 500 PAR with the LEDs higher above the tank then the T5s were.

Here are PAR values on that same tank one month after the above. This hobbyist measured 20-30% PAR drop WITH ACTIVE COOLING over the course of 6 months. After driving up the fan voltage he was able to cool the fixture enough to keep the drop to 10% over 6 months. So again, old lamps on a low quality fixture (without active cooling) will show 30% + PAR drop over the course of 15 months.

http://i43.photobucket.com/albums/e355/kangym/FTS7-31-09-002-1.jpg


Argue all you want but however you set up my test, new bulbs or not etc. my LEDs handily outperform my T5s. That's all I am claiming.

No arguing. Just trying to ensure there's a reasonable comparison without making hugely inaccurate leaps in assumptions. As shown above, your PAR values are hardly representative of a good T5 setup. And while you claim the comparison is only to your T5s you're trying to report the values as if they're some how representative of T5s by commenting on the relative quality of your fixture and expected drop in PAR of T5 lamps.

Dr. Joshi is an engineer and has been testing all sorts of lights, especially MH, for a long time so I would consider him a good independent reviewer. Here's also a more recent review of a top of the line LED fixture, the AquaIllumination:

http://www.advancedaquarist.com/2007/10/review/view?searchterm=PAR (http://www.advancedaquarist.com/2007/10/review/)

This one clearly shoes the AI unit outperforming a 250W MH unit. Granted, the MH ballast used may not be top of the line but is probably typical of one used by many people.

I'm well aware of the article. It is quite dated. And again, the comparison is to a 250W 20,000K bulb on a crappy Coralife ballast with an uknown "polished aluminum parabolic reflector". At least you're recognizing that the parameters of the comparison aren't exactly even here.

With respect to who has the biggest PAR; really, who cares. Stoney corals appear to photosaturate between 400 and 500 PAR so having 700 or 1000 PAR is probably pointless other than for bragging rights. I doubt it will do the corals much good.

FWIW, I agree. This was only commented on because Steve got involved in the conversation and has, in the past, made a big deal about the high PAR values he got with his DIY MH setups on his SPS tanks when comparisons to T5s were made. Now that LED PAR values in a typical setup (such as yours) seem to be similar to those of "good" quality T5 setups (see pictures above) he seems to be suggesting those PAR values are now acceptable.

As for longevity and degradation of LEDS, that is fairly well established. While high powered LEDs are fairly new to our hobby they have been around for a decade and are well tested. Low power LEDs have been around far longer then that so the degradation of LEDs is well understood.

With respect to PAR? Please show me the data.

[Canadian]

so I can safly say that a good LED set up such as Ron's will blow any T5 out of the watter, most 150 watt setups, and some of the 250 watt setups.

Please see the picture above of a typical good T5 setup. Would you seriously contend it blows any T5 out of the water? The numbers for the less-than-optimal T5 setup above (I only quickly grabbed those pics because I knew they existed and didn't have to spend a great deal of time searching) are not far off those that Ron is reporting with his fixture.

if he would have gone a little tighter spacing which allowed for 40 degree optics with out spotlighting I can say you would be hard to find a MH system that would be a higher output, and it would be pointless to try as at that output level would it realy matter?

Part of the fun of DIY LED I suppose - trying to optimize spacing with optic angle with height above the water without getting unattractive spotlighting. Even the better manufactured LED fixtures such as AI produce some really ugly spotlighting when suspended at suboptimal heights.

Is this because you spent to much on your T5 set up and are haveing buyers regrets Andrew? :mrgreen:

As I've stated several times I actually want to run LEDs and have for well over a year. I'm not anti LED revolution. I'm anti hype based on fuzzy numbers and marketing. I replaced my Giesemann fixture with the ATI I have now and will likely replace the ATI with a couple AI modules in the future. No buyer's remorse. I just want the best of both worlds: efficiency and PAR (not to mention "controlability") in an attractive mass produced product. But I want to make the change based on objective data and not marketing hype.

And this isn't an issue of T5s versus the world. I generally comment on comparison threads when garage "scientists" take it upon themselves to make comparisons as if they're conducting controlled experiments (not that Ron did this or is a "garage scientist" but I didn't want others to try to make comparisons as if the two fixtures were representative examples). The comparison to T5 lighting in this thread is not representative. Had Ron compared his DIY LED fixture to a Coralife MH fixture with a one year old 20,000K MH bulb I would have made similar comments about unfair comparisons. Likewise, I think Sanjay's comparison of the older model AI module to a crappy MH setup with 20,000K lamp is hardly a fair and representative comparison.

I love that you guys are helping develop LED fixtures and ideas for those of us who don't want to go the DIY route. You'll help push manufacturers to make better hobbyist fixtures. And I love that you guys are so enthusiastic about what you're doing. What my concern is, is that some people are getting caught up in the enthusiasm (and blatant marketing hype) and losing site of objectivity. Historically, the same thing happens with new lighting technology coming to the hobby. Reports about T5 longevity and PAR were completely off base for several years until real hard data started to be compiled. The same thing is happening with LEDs to some extent right now. Fortunately, LED lighting is still in it infancy (especially for use in our hobby) and is bound to improve and at the same time some controlled quasi-experimental data will start to emerge.

[StirCrazy]

With respect to PAR? Please show me the data.

Andrew, PAR is a function of spectrum and intensity, if you spectrum shifts and your intensity stays the same the PAR changes, if the intensity changes and the spectrum stays the same the PAR changes. with the LED the spectrum is the same through out the life, but like said befor at 50000 hours they will have a 15% decrease in intensity, so the PAR will drop about 15%. With the bulbs we are using now there is a intensity drop and a spectrum cange so the PAR drop can be compounded. I'll see if I can find the article I was reading about it. It comes from the glrow light side of the industry but same concerns, PAR, spectrum, intensity. this industry has had LEDs for over 10 years already, they just didn't have enough punch for coral, but they grew tomatos pretty good aparently.

Steve

[StirCrazy]

IFWIW, I agree. This was only commented on because Steve got involved in the conversation and has, in the past, made a big deal about the high PAR values he got with his DIY MH setups on his SPS tanks when comparisons to T5s were made. Now that LED PAR values in a typical setup (such as yours) seem to be similar to those of "good" quality T5 setups (see pictures above) he seems to be suggesting those PAR values are now acceptable.



Oh Andrew, I never said they are aceptible or not aceptable, if you go back and look at 99% of my posts on the T5/MH issues it was when people were claiming T5's were more powerfull then MH, and you were one. and if you read carfullyI said many a time that you could grow SPS under a NO bulb as it was done before buy a guy on this board.

your Hi quality T5 set up you showed there, is there any distance referances? and whats the point of taking readings int he air 1/2 way between the reflector and the water? so just under the surface it looks like he has a 575 average in the well lit area. at the bottom about 280ish. the hot spot is a little higher than the numbers but the other areas are lower. (I never realized the hot spot on T5s would be that pronounced) at any rate His number were very comparable with your high end T5 with new bulbs. although we would need measurments of all the reading positions to be sure. so what happens if he adds more LEDs.. nothing, bt if he adds more LEDs to reduce his spacing then adds 40 degree optics his numbers should jump by about 50% from what I have been seeing. if he adds more and then goes 20% optics.. problem is the spacing is so tight on 20% or lower optics it would only be a reasonable cost on a 20 gal tank or less, but man would you pack some punch. I know there are some people trying the the newer bulbs that have about 3 times the output as the ones Ron is using but they don't make that strength in a Royal blue yet so it is only good for the white and they are finding it is to powerfull for any other lighting they try to give it a blue look as it just washes it out. now if you like the icewhite look to your tank, which I tend to.... they might be a awsome bulb to try. I am actualy going to buy a handfull of LEDs and build a tiny cylinder that is about 24" tall but small enough aroundI can lighting it with 2 or 3 LEDs so I can try some different combos on my Ardunio and see what kind of color blending, lenses and PAR values I can get a different depths. just have to clean my shop out first.. which at the way I am goiong should be another 6 months :redface:

Steve

[Canadian]

Andrew, PAR is a function of spectrum and intensity, if you spectrum shifts and your intensity stays the same the PAR changes, if the intensity changes and the spectrum stays the same the PAR changes. with the LED the spectrum is the same through out the life, but like said befor at 50000 hours they will have a 15% decrease in intensity, so the PAR will drop about 15%. With the bulbs we are using now there is a intensity drop and a spectrum cange so the PAR drop can be compounded. I'll see if I can find the article I was reading about it. It comes from the glrow light side of the industry but same concerns, PAR, spectrum, intensity. this industry has had LEDs for over 10 years already, they just didn't have enough punch for coral, but they grew tomatos pretty good aparently.

Steve

And you're assuming that with whatever installation method you choose for heatsinks and whatever you choose to drive the emitter at there will be negligible thermal damage. Again, and again I say this: theory is great when you don't have real data. We can assume there will be little intensity drop but until we see long term data, especially for the blue emitters that are more prevalent in our hobby, it's just an assumption.

[Canadian]

Oh Andrew, I never said they are aceptible or not aceptable, if you go back and look at 99% of my posts on the T5/MH issues it was when people were claiming T5's were more powerfull then MH, and you were one. and if you read carfullyI said many a time that you could grow SPS under a NO bulb as it was done before buy a guy on this board.

your Hi quality T5 set up you showed there, is there any distance referances? and whats the point of taking readings int he air 1/2 way between the reflector and the water? so just under the surface it looks like he has a 575 average in the well lit area. at the bottom about 280ish. the hot spot is a little higher than the numbers but the other areas are lower. (I never realized the hot spot on T5s would be that pronounced) at any rate His number were very comparable with your high end T5 with new bulbs. although we would need measurments of all the reading positions to be sure. so what happens if he adds more LEDs.. nothing, bt if he adds more LEDs to reduce his spacing then adds 40 degree optics his numbers should jump by about 50% from what I have been seeing. if he adds more and then goes 20% optics.. problem is the spacing is so tight on 20% or lower optics it would only be a reasonable cost on a 20 gal tank or less, but man would you pack some punch. I know there are some people trying the the newer bulbs that have about 3 times the output as the ones Ron is using but they don't make that strength in a Royal blue yet so it is only good for the white and they are finding it is to powerfull for any other lighting they try to give it a blue look as it just washes it out. now if you like the icewhite look to your tank, which I tend to.... they might be a awsome bulb to try. I am actualy going to buy a handfull of LEDs and build a tiny cylinder that is about 24" tall but small enough aroundI can lighting it with 2 or 3 LEDs so I can try some different combos on my Ardunio and see what kind of color blending, lenses and PAR values I can get a different depths. just have to clean my shop out first.. which at the way I am goiong should be another 6 months :redface:

Steve

False. Steve, I never once said that T5 was more powerful than MH. That is patently false. I expect you to go back and provide a quote in context to substantiate that I ever said it was more powerful. All I ever maintained was that T5 was not inferior to MH when you made broad-based statements about inferiority and superiority one way or the other.

The tank pictured above is 5' with a 4' fixture over it. The tank is 60 x 30 x 24. The fixture is roughly 6-7" above the water's surface. The edges of the tank are dim because the fixture does not extend to the edge and because the first 3" or more of a T5 lamp produce significantly less light. Within the 3.5" feet centered below the fixture the PAR is pretty consistent (biased on one side because that's where the fans are).

[Canadian]

And Ron, I apologize for derailing your build thread.

[OceanicCorals-Ian-]

I have a 48" 10 bulb ATI Powermodule on my home tank, the 48" Aqua Illumination module I have seen is brighter than my Powermodule. This of course is just my observation.

On a second note, Ron, you are more than welcome to come over and pull some numbers off my tank. Also, it would be nice to see all of this back and forth discussion on the comparatives between LED and conventional lighting moved to a new thread for this purpose. I was actually just enjoying Ron's build thread as it was.

Ian

[OceanicCorals-Ian-]

I have created a thread specifically for this type of conversation, if everyone is interested maybe the MODS can move the latter half of this vs conversation to the thread I have linked below.

I feel this is detracting from this excellent build thread......

http://www.canreef.com/vbulletin/showthread.php?p=510374#post510374

[Ron99]

I have created a thread specifically for this type of conversation, if everyone is interested maybe the MODS can move the latter half of this vs conversation to the thread I have linked below.

I feel this is detracting from this excellent build thread......

http://www.canreef.com/vbulletin/showthread.php?p=510374#post510374

Thanks Ian. Good Idea. Let's move the debate over there.

And Ron, I apologize for derailing your build thread.

No Worries, it's good to discuss it and try to develop real information. I agree that there is lots of manufacturer hype and BS. One reason I built my own array, aside from costs, is that most of the commercial LED lights make compromises of some sort unless you spend big dollars on the high end fixtures.

In any case, let's continue in the other thread.

Where did you find most of your reading and research?

nano-reef.com seems to have the most active LED community but more is showing up on reef central now too.

[BKTruong]

Hey Ron great thread I've been following since day 1 but I want to know now that you have your new led fixture up for a few days running, how are all the SPS corals doing? Also do you find that your LEDs miss some light in some areas as your LEDs are spreaded out and LED light rays are directed in a straight focused downward "UFO" path and doesn't light its surrounding area like MH or even T5s. As you may have probably already noticed, the solaris's or any professionally made LED fixtures are all built with its LEDs side by side tightly snug. I honestly prefer to spread them out as you did as it covers more area but wasn't sure if it was ideal.

[DiverDude]

<snip>One reason I built my own array, aside from costs, is that most of the commercial LED lights make compromises of some sort unless you spend big dollars on the high end fixtures.
</snip>
Y'know....sometimes, sometimes it's just fun to build your own stuff. You get to make it suit your personal needs exactly and there is a great deal of satisfaction to be had from the design and build process.

After all, this is a DIY forum, is it not ?

[byee]

Totally agrees..........GREATER satisfaction from it plus you the educational aspect.

Probably built better as well. :lol:

[Acrylic]

This should shed some light on the subject. (pun intended :) )
http://www.cree.com/products/pdf/XLampXR-E_lumen_maintenance.pdf

[Ron99]

This should shed some light on the subject. (pun intended :) )
http://www.cree.com/products/pdf/XLampXR-E_lumen_maintenance.pdf

Pretty much what we have been discussing on the other thread. Keep the LEDs cool and they will last for a long time.

[Ron99]

Time for a quick update. Haven't really had time to work on the controller very much but hopefully soon. But after about 1 month I thought I would post a couple of pictures of the corals.

Here's a shot taken on April 23rd:

http://hotimg23.fotki.com/a/70_70/254_23/Coral11.jpg

And one from May 19th:

http://hotimg23.fotki.com/a/70_70/254_23/Coral12.jpg

It has great colour and some noticeable growth in the last month.

Here's one that was pretty much brown except for the tip:

http://hotimg23.fotki.com/a/70_70/254_23/Coral21.jpg

Now it has a bit of growth but is colouring up all over:

http://hotimg23.fotki.com/a/70_70/254_23/Coral22.jpg

[Dyspnea]

Looks great!

What was your lighting setup before the LEDS?

[aquajeep]

curious as to what optics to use?
http://www.ledil.fi/datasheets/DataSheet_LC1-TAPE.pdf or http://www.ledil.fi/datasheets/DataSheet_CRS.pdf
they have real spot or smooth spot versions too.as for drivers i was going to go with LM3464 from national semiconductor.
( i got data sheets from Walter Shawlee 2, President - Sphere Research Corporation
3394 Sunnyside Road, Kelowna, B.C., CANADA V1Z 2V4
URL: http://www.sphere.bc.ca E-Mail: walter2@sphere.bc.ca )

[Ron99]

curious as to what optics to use?
http://www.ledil.fi/datasheets/DataSheet_LC1-TAPE.pdf or http://www.ledil.fi/datasheets/DataSheet_CRS.pdf
they have real spot or smooth spot versions too.as for drivers i was going to go with LM3464 from national semiconductor.
( i got data sheets from Walter Shawlee 2, President - Sphere Research Corporation
3394 Sunnyside Road, Kelowna, B.C., CANADA V1Z 2V4
URL: http://www.sphere.bc.ca E-Mail: walter2@sphere.bc.ca )

I don't think any of those optics will be best as they seem to be very narrow (3 degrees to 21 degrees). Depending on what you want to do you should be looking at 40 degree or 60 degree optics. Something like the ones nanotuners sells. I don't know the brand off hand but they are the type that most DIY folks are using.

As for the drivers, they seem a bit complicated to implement. I would look at the Meanwell drivers as they are far simpler. They run off AC power directly and you can get the dimmable versions. So the only small added complication is circuitry for dimming input but that can be fairly simple too if you just want a knob to dim them.

[globaldesigns]

I have personally seen Fishytime's lighting, and it looks fine with 2, put together.

[aquajeep]

in my email i got from him he said they had 54 and 62 degree optics.but there are diff kinds of lens ,real spot, oval,rectangular ect ill have to see if he can get the meanwells in.prob cant thats why he sent me the other info im guessing.what prices were you able to get your stuff for?

[Ron99]

in my email i got from him he said they had 54 and 62 degree optics.but there are diff kinds of lens ,real spot, oval,rectangular ect ill have to see if he can get the meanwells in.prob cant thats why he sent me the other info im guessing.what prices were you able to get your stuff for?

Sorry for taking a while to answer, been a bit busy lately. I bought most of my parts through a group buy at nano-reef so the prices were good. Star mounted LEDs were $5.75 each and optics were another $1 IIRC. Nanotuners has a good sale on now on LEDs, optics and drivers.

[Twinn]

Great build I feel like I'm late to the dance on this one. I think a real big point that everyone is missing is the electrical consumption used to power leds vs T5 or MH. Anyways any updated photo's or info on the controller.

thanks

[blkhwk]

Ditto on that Twinn and less heat too..Just wanted to add excelllent job on the light and the post Ron..hats off to you

[Bblinks]

SORRY TO HEAR THAT YOU ARE SELLING YOUR LIGHTING, AND OF COURSE THE SITUATION AT HOME. :cry: