The budget bruiser LED build...

[lingwendil]

i'll need to really figure out what i'm looking at as far as salt spectrum now that we figured out how to add COBs to spectra. this is what I kinda have in mind for a setup for my 20cube, but i'm not sure what i'm really looking at but I guess its somewhat cheap to tweak down the road.

Crisp White- Star (Cool Blue Cyan Lime) - K16 - Star (cool Blue Cyan Lime) - Crisp White. maybe add a couple violets in as well?

this will be mounted in a line like that tight together heavy whle I want the 2 crisp whites even if I run them dimmer just so its an even spread for the most part. or maybe run 2 k16s on the outside and run them dim since there cheap anyway and the white in the middle.

planning on using 1ft steve's HD heatsink since it can be run more or less passive from what i read at 1 ft for 50-60 watts of power which I doubt i would go over when over a 35 gallon cube.

I think that layout with an added pair of U60 or U50 Violet (or one of each, one on either end) from RapidLED would be a killer setup that covers a good range of spectrum pretty well. If you punch everything into Spectra and look at the advisor it gives a "100" rating from everything before you even add the royal blue at all once you add one each of the U50 and U60 violets in there. You get a nice heavy violet peak, followed by a heavy broad-range blue, and then a good flat line out into the red region. I would think it would be a pretty good setup for sure.

Just for spread and color mix it wouldn't hurt to add another K16, and have one on either side of the strip, flanked by a white and the lime/cyan/blue 3-ups, with an additional violet added too, just means another driver, which if using LDD-H series, is no big deal. Gives you the opportunity to run them at a lower duty cycle and get a nice long life out of them too.


What controller do you plan on using? That can make a difference too. It would be a shame to not break out each color onto its own channel for such a rig. I have one of these running my 20 long, and it can handle up to sixteen (!) channels, and I currently run six time slots if I'm not mistaken-

http://www.ultimatereef.net/threads/simple-multichannel-up-to-16-led-controller.784531/

Best of all its CHEAP! Same cost as building a Typhon clone really, and very versatile, perfect for a power user that doesn't need a ton of knobs and doohickeys all over.

 

[kampo]

Leaning toward using my sslac16 controller o2 made up for me. Wish I had him do more up for me. For what I paid it can't be beat.

I like the idea of what your saying. And honestly all the colors are there and way brighter than I need so if the rationis off I have headroom to adjust it in software.

 

[kampo]

eBay had a 15 off 75 coupon tonight so snagged 2 makers 12in slim heatsinks, k16 chip and a tc420.

So going to use the tc420 controller. Not ideal, but it should work fine. So I need to turn this 6 channel array into 5

Currently plan was
Ch1 k16
Ch2 2x 1202 crispwhites
Ch3 2x cool blue
Ch4 2x cyan
Ch5 2x lime
Ch6 2x hypervilolet

Thinking just combine 6 and 1?

Also 12in makers heatsink. You guys think I'll need a fan on it? Saw another person use a pci cooling fan inside the heatsink which was pretty slick.

 

[oreo57]

eBay had a 15 off 75 coupon tonight so snagged 2 makers 12in slim heatsinks, k16 chip and a tc420.

So going to use the tc420 controller. Not ideal, but it should work fine. So I need to turn this 6 channel array into 5

Currently plan was
Ch1 k16
Ch2 2x 1202 crispwhites
Ch3 2x cool blue
Ch4 2x cyan
Ch5 2x lime
Ch6 2x hypervilolet

Thinking just combine 6 and 1?

Also 12in makers heatsink. You guys think I'll need a fan on it? Saw another person use a pci cooling fan inside the heatsink which was pretty slick.

how about 4 and 5?

Oh, and here is lime with a little cyan mixed in.

 

[kampo]

that was my second choice, actually considered just dropping the channels to 4 to save a bit on drivers, leaning toward using 2x Blue Acro drivers, didn't really want to buy 3 of them and waste a channel. that said I could buy 3 and just combine the control on 2 channels but still have control over all 6 because his drivers are adjustable...hmmm

also why is there no edit option for threads for me I see you have that option?

 

[oreo57]

that was my second choice, actually considered just dropping the channels to 4 to save a bit on drivers, leaning toward using 2x Blue Acro drivers, didn't really want to buy 3 of them and waste a channel. that said I could buy 3 and just combine the control on 2 channels but still have control over all 6 because his drivers are adjustable...hmmm

also why is there no edit option for threads for me I see you have that option?

Time to edit.. "times out"....

Never timed it.. LOL..

 

[kampo]

lingwendil have you put together your full array yet and have thoughts? just curious of the brightness of each led to see if my version of it will have enough oomph in the colors.

 

[lingwendil]

Not yet. Been waiting on spare time, and working my life away. Coincidentally, just this morning I had to take my commuter car (73 super beetle) off the road due to severe engine issues, and need to spend my spare time and cash this week on tossing another running motor in there...

I've got a line on some PCBs to simplify things, but ideally an 8x driver board would be best, but I suppose a pair of 4x LDD boards would work.

 

[Maryland Guppy]

, but ideally an 8x driver board would be best,

http://www.stevesleds.com/LDD-Driver-Array--8X-Drivers_p_306.html

 

[kampo]

http://www.stevesleds.com/LDD-Driver-Array--8X-Drivers_p_306.html

i'm kinda leaning toard using blue acros drivers. i like that you can tune the output indepedently of the pwm. so you can actually tune in your spectrum there, and just use the pwm for sunrise/sunset. cost is actually cheapper than using a board + LDD drivers for the most part too it seems.

 

[Gorgok]

i'm kinda leaning toard using blue acros drivers. i like that you can tune the output indepedently of the pwm. so you can actually tune in your spectrum there, and just use the pwm for sunrise/sunset. cost is actually cheapper than using a board + LDD drivers for the most part too it seems.

Tuning for spectrum and still running sunrise/set with pwm is trivial in software. Basically you just tune the thing when its running 'full' and then sunrise/set does a simple % math on that.

 

[MrMan]

Tuning for spectrum and still running sunrise/set with pwm is trivial in software. Basically you just tune the thing when its running 'full' and then sunrise/set does a simple % math on that.

By tuning the current you keep the full pwm resolution allowing for smoother ramping. If you use PWM to adjust the LEDs to run at 40% and then dim down to 0% you're only getting 40% resolution on dimming. Versus dimming current down to 40% and then PWM from 100 to 0%.

Not really an issue with 12-bit dimming but if you were using an arduino with only 8-bit it could become noticeable. Plus LEDs are more efficient at lower currents aren't they?

 

[kampo]

By tuning the current you keep the full pwm resolution allowing for smoother ramping. If you use PWM to adjust the LEDs to run at 40% and then dim down to 0% you're only getting 40% resolution on dimming. Versus dimming current down to 40% and then PWM from 100 to 0%.

Not really an issue with 12-bit dimming but if you were using an arduino with only 8-bit it could become noticeable. Plus LEDs are more efficient at lower currents aren't they?

pretty much nailed it. for instance if your never going to use more than 50% of your whites, no point in sending them 50% PWM at 1 amp, mine as well send them 100% at 500ma, the latter is much more efficient and the leds will run cooler.

 

[Zephrant]

Not sure I buy the cooler at 100% PWM and 50% amps. It would be dissipating the same wattage either way.

Or are the LEDs more efficient at 50% amperage? In that case, not cooler, but more light for the same power.

 

[kampo]

so got my heatsinks today, really nice super excited to get the rest on the way, been super busy at work so haven't had a chance to decide on my final layout. thinking i'm going to do this in phases i'm new to reef so not completly sure what way I want to go yet as far as coral stocking ultimately anyway. thinking phase one this is going to be my layout

2 channels:

Channel 1 2xk16 2xhyper Violets in 2 chains run in parrell at prolly 700ma
Channel 2 leaning toard just getting a trust vero decor since they run at a lower voltage, and will let me then get 2 stars of (cool blue, lime and cyan) on the same channel.

that way I can save money on the driver side for now and keep it simple. phase 2 will prolly be to snag another 2 channel driver from blue acro and ad 2 more violets and put them on there own channel, and tweak the colors and divide them up a bit better and go to 4 channels.

is my plan of running the blue violet together in a parallel channel horrible? they will both be running under spec amp wise so figure should be ok.

 

[kampo]

Not sure I buy the cooler at 100% PWM and 50% amps. It would be dissipating the same wattage either way.

Or are the LEDs more efficient at 50% amperage? In that case, not cooler, but more light for the same power.

you get more lumens per watt at lower amperage. so more of that wattage is
being converted to light not heat. also another thing to think of. if you have
4 chains of leds running at 48V at 1 amp you would need a 200 watt PSU if its running at 100% or 10%, but if you have that same string and only really need 50% you can run it at 500mah and only need a 100 watt psu.

 

[lingwendil]

so got my heatsinks today, really nice super excited to get the rest on the way, been super busy at work so haven't had a chance to decide on my final layout. thinking i'm going to do this in phases i'm new to reef so not completly sure what way I want to go yet as far as coral stocking ultimately anyway. thinking phase one this is going to be my layout

2 channels:

Channel 1 2xk16 2xhyper Violets in 2 chains run in parrell at prolly 700ma
Channel 2 leaning toard just getting a trust vero decor since they run at a lower voltage, and will let me then get 2 stars of (cool blue, lime and cyan) on the same channel.

that way I can save money on the driver side for now and keep it simple. phase 2 will prolly be to snag another 2 channel driver from blue acro and ad 2 more violets and put them on there own channel, and tweak the colors and divide them up a bit better and go to 4 channels.

is my plan of running the blue violet together in a parallel channel horrible? they will both be running under spec amp wise so figure should be ok.

Eh. I'm not a fan of running arrays in parallel without some way to guarantee balance. I wouldn't.

 

[biecacka]

Interesting read. Not much knowledge of diy led here, but I'm tagging along. If I ever where to switch back to led, I would look into something like this possibly. A large panel similar to the orphek or something to light my tank.

Corey

 

[kampo]

lingwendil did you ever get your light built? i'm hopefully going to assemble my light this weekend. but won't even start cycleing the reef tank till early next week so it will be awhile till I'm lighting anything coral like lol.

 

[ReefkeeperZ]

that was my second choice, actually considered just dropping the channels to 4 to save a bit on drivers, leaning toward using 2x Blue Acro drivers, didn't really want to buy 3 of them and waste a channel. that said I could buy 3 and just combine the control on 2 channels but still have control over all 6 because his drivers are adjustable...hmmm

also why is there no edit option for threads for me I see you have that option?

I asked the same thing in another thread, it's based on post count (over 50) and duration I believe they said it's 90 days