Jestersix

Gomer's LED array

Hey Tony, I have a couple of sm. tanks at the shop that I need some advice on how to light by LED. LMK if there's a time we can chat about it. Thx-Jim
 
I decided to put the $ into perspective so lets figure out what this guy costs compared to MH and what is the time period for payback. I'll just figure it out as I type this post out :p



Old setup was a 250watt icecap with radium bulbs changed every 6 months and LumenMax reflector.

Icecap Balast = $130
LumenMax 2 = $154 (but I paid $100 used so lets use that number)
Radium Bulb = $60 each (this is cheap..I think I paid 80 or so for my first one. I think $60 was the group buy price)
Light consumes 262 watts.


LED cost me $325 or so in parts. $300 in GB stuff and I added $25 for some misc items I picked up. Probably not that much, but oh well
Power: LED is ~130 watts. driver is ~88% efficient, so that is 148watts

Power: I ran 7hrs a day. I'm Teir 1 electricity @~0.121/kwh

oimg
 
I think your cost payback time is a bit silly because from day zero it was cheaper :D

That's the type of cost analysis that I often see with the LED setups... however lets flip the script a bit, instead of factoring in the cost of all the MH equipment, lets say you had the ballast, you have the reflectors, and for kicks and giggles lets say you need to replace your bulb, so the initial cost of your MH is $60 not the $300+ you start it at. Also since your graph is linear I don't know at what interval that you consider the time to replace your bulbs at.
 
It's relevant from the stand point of those with an option now.
If you want to assume any initial starting conditions, it is merely an offset.
The plots are linear because I chose datapoints at 6month intervals which is the interval that I replaced radiums.
Old setup was a 250watt icecap with radium bulbs changed every 6 months and LumenMax reflector.
 
but to humor you, here are the plots again

1) if you had to chose a new fixture
oimg


2) if you compare MH with new bulb and upkeep vs new fixture and upkeep
oimg
 
Let's just say you had the MH up and running and your only cost was bulb replacement. For the sake of argument or laziness, let's say that you only changed those bulbs once a year. At the same time, you want to compare how much it would cost to crank out a DIY led setup. Let's just say for kicks and giggles, it costs you $750 (double Tony's figures above) for all the parts necessary to set it up.

MH setup (2 bulbs) = 150/yr.
Bulb replacement = yearly

LED setup = $750 (72 led array)
LED life = 50,000 hours
Run time = 10 hours daily (13.7 years)

You'll break even within 5 years. This doesn't even take into account the cost savings in electricity.
 
Perhaps true.... two big issues I have with this

1) You're stuck with that exact same setup of color/intensity, I don't know about you but I've gone through more bulbs colors than I care to think about.

2) You're making a huge assumption on rated lifetime of those LEDs... I've seen enough dead LEDs in stoplights to know they can potentially go out early. Especially considering the environment they're in over our tanks.

I love the idea behind LEDs, I'm not trying to bash Tony's project here, I think it's awesome. However they're really at the stage where computers were about 10 years ago, where they are nice an economically viable if you go the DIY route. Not everyone can solder wires to a circuit board, and if you buy one of the commercial fixtures that can easily be 2-3 times the cost of Tony's for an equivalent amount of watts then your buy back time goes through the roof. But I guess you can make the argument for any fixture, however a retro MH or T5 fixture is a helluva lot easier to throw together than an LED fixture ;)


Also has there ever been any long term (1 year? 2 year??) studies (amateur or what not) to show that you can grow corals just as successfully?
 
1) if you wire it up as such, you can change the ratio of "white" to blue very easily. You can also, with very little investment, add in a couple additional different colored LEDs (red, green, amber etc) to add a new hue if you really feel like it.

2) LED stoplights are a topic of discussion on LED forums. Poor implementation is their downfall from all I gather. My heatsink at 1amp is just slightly warm. I hate to think of how hot the high density LED boards in stop lights are. I also way underated the lifetime of my LEDs. My graph stops at 5 years. I only gave it 7hrs a day. That is only ~13k hrs out of their "predicted" 70K hour lifetime :)

I don't think it is prime time for commercial units. Not economic. I do think the DIYers can get in the market now.

I'm not sure of long term LED Reef studies (but there are likely some solaris units that are that old). There are long term studies on plants. If the concern is the spectrum of LEDs, then I'd challenge it with "where are the 1-2 year studies with the new MH/T5 bulb by ____" They are technologies to give photons. They all can give the fux. MH/T5 has so many spectral variations it isn't even funny. No one challenges they even when new specral variations come out ;)
 
LEDs can be dimmed and henceforth you can change the spectrum somewhat. I do know about the swapping through bulb colors myself and still have 7 extra 150W bulbs of various colors that I've purchased over time to see what each looks like. Here are some pics to illustrate (28 LED array; with 2 reds throw in with the blues and whites).

IMG_0045.jpg


IMG_0046.jpg


IMG_0047.jpg


IMG_0048.jpg


IMG_0051.jpg


IMG_0052.jpg


As for longevity going out to 1-2 years, not that I know of. I've only seen 9 months out report and yes, the corals are still growing.
 
Well Tony the reason why I questioned the LED growth is very spectrum dependent because with MH bulbs (and to a minor extent T5s) you have a continuous spectrum (albeit very low temp) with an emission spectrum tacked on on top of that.

From what I remember of LEDs is that if you have a 420nm LED you basically get 420nm of peak, with very little on either side of it, and absolutely none any significant wavelength away. I'm not quite sure how the white LEDs compare to that, but I'm going off older (more physics based :D) knowledge.
 
Mike, the LED spectrum is also continuous. "white" isn't an RGB discrete mix. It is a phosor coated blue LED I'll give it to you though, that LEDs lack the 420nm that some MH do well. T5's not so much.

As for moisture, my LEDs are way above the water level. I can easily get my head between the LED and water level to look around etc. 18" + above. No worries for me :)

33858156.whiteledspectrum.jpg

fig10-iwasaki6500K.gif
 
hey gomer,

whats your thermal junction temp at? And how are you measuring it? Mine seems to be on the high side and im a bit worried =/
 
Not measuring it. All I can tell you is that the MCPCB is the same temp as the heatsink (to touch) and that temp is barely warm.
 
oh? my pcb kind of actually burns if I keep my finger on too long but that might actually be the lense. It's cool to the touch if Im just doing cursory touching. The heat sink is cool as well but when measuring the leds, I got 195 at the top of the lense and 120 as I drop down the radius.

Are you at all worried about the heat issue or is that kind of a nonconcern? Cree's spec sheet says maximum TJ of 150C. Im not sure if I should be using the temperature on the top of the lense (not optic) or the solder point next to the emitter in the calculation.
 
Back
Top