Thanks This all started as me just posting my findings way back when when I started experimenting with high power LEDs over a pico that I built. That experiment just grew and grew. Fortunately, we have a good community here that started to bring up a lot of questions that I started to research and report on. It snowballed from there. I have always been about bettering the reef community with this technology and trying to educate everyone about their potential. LEDs got a bad rap early on when little was known about their potential, and I've been trying to correct that.

We still have a lot to learn about them. I have some theories that I'm looking to try out. Hopefully we can all push LEDs into the mainstream in the not too distant future.

is there any difference in using the red computer motherboard washers compared to the nylon to secure the starboards

Nope. They are phenolic, and are used as an electrical insulator also. If you have a bunch, use them. They are a little more brittle though.

perfect just wanted to make sure

just need a quick confirmation... i bought a 1Ohm resistor the other day and hooked it up in SERIES with my led array after the led's... i measured VDC at each end and the resistor measured 1.52VDC.. does that mean im running my LED's at 1500mA?? how the heck can i reduce that.. im using the ELN-60-48 with 11 led's currently...total volts used over the whole array is only 39.8 or something close to that..

if my maths is correct 1 led (3.7v @ 1amp) times 11 = 40.7.. what am i doing wrong.. (my multimeter does not read current for some reason.. )

what else can i check to see if im over runnign these led's.. is it even possible to run them at 1.5 amps with this driver (1.3amps max)

If you can safely open your ELN-60-48, you will see a screw which you could turn and bring down to just under max 1Amp. Your current setting inside the ELN is set to max, screw it down to desired requirement. I hope this would help....

that was the original problem.. when i turn the screw nothing happens, not even for the voltage screw changes when being turned..

funny thing though.. with the current screw if i push down on it (not heavily) the voltage across the resistor changes to 0.89# something.. but as soon as i relive pressure it goes back up to 1.5##

That does sound odd. Sounds like the pot may be bad if you can't adjust it and the voltage across the resistor changes with pressure. You are measuring everything correctly though. As a stop gap measure, you could put a little lump of foam on top of the pot so that when you put the case back together it applies enough pressure to it to get it to work.

HMMM, i prefer to kep the cover off because of how hot it gets. i think it needs a small heatsink mounted ontop of the alluminium chuck it has there..

when there is downward pressure applied, the voltage drops to 0.890 (890mA)...

if i keep them cool, how long do you reckon the leds will last at 1500mA (if that's what it actually is..)

Hard to say, but probably reduce the life to under 10000 hours.

hmmm... dang it.. can i add a few led's so it drains more or something.. this sucks..

Hi,

Does anybody have an idea of the input impedance on the 0-10V input on the Mean Well ELN-60-48D?
I will put 5 in parallel for the RoyalBlue's and the same for White, and control them by a GHL,
but i want to add a serial Pot so in can adjust the max output current by lovering the control voltage.

Should i put it in series like this?
GHL -> Pot -> 10v input on meanwells.

Or connect the ends of the Pot to the +- on the 0-10V output and connect the middle to the Meanwells?

I want to have the full resolution on the GHL, but be able to turn the max output down to around 700mA. so i by adjusting the Pot can adjust the balance of the leds without reprogramming the GHL.

Does it make sense?

There isn't a setting in the Profilux for the max output? On the RKE/RKL, you can set the max, then your ramp up/down is 0-100, with 100 being your set limit.

Maybe, haven't bought it yet, waiting for my Meanwells.
On the RKL, if you set max to e.g. 5V do you still have 100 levels on the ramp, giving increments of 50mV?
I will look in the online documentation for the GHL.

That seems to be the way it works. At least, that's the way it was demonstrated to me at MACNA by a DA rep. I'm sure there are limitations on the hardware on how fine the resolution is.

Excuse me if I'm not understand terminology and jumping to conclusions but did I just read that the RKL has the ability to control the LEDs similar to these costum controls with ramping blues up first, then whites a little later, then ramp whites down and then ramp blues down? I know it could control the powering on and off like any wall timer. I'm probably just not understand you though.

I don't think it can do seperate ramp up/down of the two channels, but you can set the max point for each channel, basically allowing you to set your color temp. Then you can set the amount of time that the channels take to transition, but I think they are done together. I'll have to hook up my RKE tonight and check the programming (don't have an ALC yet to check the actual output)

Hi Evil66,

I've read most of your LED posts here and on other forums, you have a great amount of knowledge on the subject and take a lot of time helping others. I'm sure not many say thanks, but we all owe you a huge debt. THANKS

You're certainly welcome

just a quick question.. why dont more people just use resistors instead expensive drivers.. i understand they add SUBSTANTIAL amounts of flexibility compared to resistors, what are some other reasons.. i was thinking about doing 2x3 led series over my FW tank just for fun, but i dnt want to fork out 30-40 dollars for a buckpuck.. so i was just going to use a smallish DC power supply and some small resistors.. and it will save me like 30 bucks, am i missing something here..

ALSO can i glue the resistors to the heatsink.. to help dissipate heat from them or is there a law against that?

ONE MORE! i know you always talk about leds failing open and close and stuff, but what are the odds that the led actually fails.. i have abused one of mine making flashlights and doing little tests on it and it still doesnt die..

also just in case you were wandering why buckpucks are so expensive, i live in Australia...

Resistors can work, but they offer no protection. They don't protect against thermal runaway or changes in input voltage. They are also horribly inefficient, and cannot adapt to variations in input voltage. Considering how much the LEDs cost, I'd think that you would want to protect them. Anyway, they offer a much easier way to dim the LEDs.

hmm.. im gonna risk it, im going to drop off 6 white leds (off my array) cause im swapping with blues so i have a few leftover...

now the problem i have is that 2 online calculators give me 2 variations of answers... the first tells me that i can run 2 parallel series of 3 leds at 1AMP with a single 1 Ohm (1watt) resistor before each series and be done with it..

but the next online calc says that if i do that then it will be concerned about the power dissipated in my resistors.. what will happen, heat up and melt?? blow up..

ALSO they both tell me that the array will draw 2000mA from the source... does that mean that my PS 12vdc @ 2.2Amps is more than enough, or does it mean i have to put a resistor before the array to bring it down to 2Amps??? or will it just draw what it needs?

Need a little help. I just finished my DIY light and my blues work but the whites dont. Here are the details:
Cree LED's Q5's and RB
Meanwells 60-48D's
I have the meanwell dim wires hooked up to a Reef Kepper light ALC that dims the lights 0-10v. The blues work great but the whites just have a really small amount of light coming from them, a small glow. Then when I unplug the power the blues go out and the whites flash really bright for a split second and go out. Any things that I should be looking for before I go through every bit of wiring? Thanks for any ideas.

OK I am an idiot I found my problem. I wired the switch from the RKL ALC backwards, all good now works great and thats the only problem I had with the wiring. I thought I would have at least 5 or 6 fixs I would have to do considering I had to do over 330 solder joints and many wire extensions!!! I have to thank Evil as well cause the only reason I took on this project is because of his research and info he gave me, so thanks buddy. Enjoy the pics:

Wow, that's alot of LEDs!

If you are running two parallel strings of 3 LEDs on a 12v power supply, you will need a 1ohm, 1W resistor on each string. Your current draw will be about 2A, because that is what you are setting the draw to with the resistors. I'd still advise against this though. If you are looking to go cheap on the current regulation, there are better alternatives to resistors that are still cheap.


D@mn that's a lot of LEDs!! What was this going over again? Good to see it was easy to get working with the ALC. I need to order one of those and start playing with it.

What purpose is the ALC serving in you're build? I've been considering purchasing a RKL and have been wondering what I could get it to do for my lights.

The ALC will allow you to set your white and blue intensities, and ramp up and down during your on/off cycles, so you get a soft on and soft off. That's about the limits of it's control capabilities so far. DA may add some more flexibility in future firmware updates, if they are feeling generous

hmmm. ive done alot of searching lately and only other thing ifound was diy driver...

please explain to me how bad this can be..

i 6 leds will need zero flexibility as it will only be turned on an off

The biggest issue with resistors is that they are not smart. Setting the current with a resistor relies on two things; the voltage drop accross the LEDs, and the input voltage from the power supply. Both have to be rock solid steady for a resistor to work perfectly. Neither will be though.

The voltage drop across the LED changes as a result of heat, and will fluctuate slightly during it's operation. The power supply can also fluctuate slightly based on changes in voltage at the wall, which you can't control. An example would be something like this:

If the voltage is12v from the power supply, 1ohm 1W resistor in series, and the LED voltage drop on each LED drops to 3.5v, you are now running at 1.5A. While that's fine for a while, the additional heat causes the voltage to drop even more, increasing the current. This is thermal runaway, and it will kill LEDs. It's an exagerated example, but it happens.

Even a simple transistor based linear regulator will keep the current in check better than a resistor.

Bigrock, are you planning on cooking rotisserie chickens under that array!? That's 140 LEDs!

or using them for illegal activities? just remember, separate the males from the females.

Unfourtantly this is just costing me money and will not have a return on investment like you are impling:)

This light is going over a 120g 48"x24"x24". I will be using 60 degree optics on all but maybe 20 where Ill have 40 degree optics. I wanted a uniform light where I wouldnt have to worry where I placed corals and it still cost alot less than the manufacured lights.

The chicken Idea is interesting though maybe you are on to something. Ive heard of infared BBQ's but not LED BBQ's:)

Very cool. Make sure you post lots of pictures once you get it up and running over the tank.

Evil, would you say that a good starting point for the number of LEDs we should run is 4 LEDs per 12 inches of tank length and 1 row for every 6 inches of tank width?

I think it would be more accurate to say one row per 3 inches leaving a 3 inch gap all the way around. As for the 4 per 12 inches thats dependent on depth and livestock choices because of optic requirements.

I know I'm not evil, but this was my understanding and how I did mine.

do any of you guys happen to know if a vga fan blows up or out the sides?

also, a few other technical questions i had about heatsinking are as followed:

1. minimum surface thickness of heatsinks
2. minimum space between fan and heatsink surface
3. minimum thickness of heatsink wings
4. minimum spacing between heatsink wings
5. minimum thickness of back material to use bulk diffusion and not have to cnc in heatsink wings

the questions are for lets say... 6 emitters/12emitters for example on a 3x9 area

It's difficult to answer, as every tank has different requirements. The basics I use for figuring out what is required, is to take the length of the tank, subtract 4-6" (depending on the size of the tank), divide by 2 and that gets you the number of LEDs per row. For the number of rows, I normally go 2 rows up to 12" wide(unless it's a cube, and I normally go up to 3), 3 rows up to 16", 4 rows up to 18", and 5 rows up to 24". Again this is all dependant on what the requirements are of the tank. Tighter optics will force more rows.



1. More surface area the better. Anything more than the surface area of a flat sheet is better than a flat sheet
2. Depends on the fan type. Axial fans should have a gap between it and the base. It helps reduce dead spots under the fan. Maybe 1/4-3/8" minimum gap? Radial fans obviously can be mounted right to the base.
3. Depends on how well you can machine it. Thin fins pull heat away faster, but heatsoak quicker.
4. Fins should be spaced apart 1.5 times the fin thickness. It helps reduce the self heating effect (one fin heats another, which heats another, etc...)
5. Minimum thickness should be determined by the fastener used to bolt the LEDs down. Typically you want 2x the thread diameter as a minimum threading surface thickness.

ugh i know i keep bugging you with PMs with me building my own vs what your offering for sale....

i think after sending that last PM and reading this for the 6th time, i might try it on my own. i have a 3 gallon JBJ picotope.....and i think i may build my own array.

odd question here, but cam i make a 1 or 2 LED mini fixture to sit on my millenium 2000 (refugium)?

[_main tank__]--[_fuge_]------Power. something along those lines?

also if all your LED monsters cna help me with a shopping list id really appreciate it. ive seen that cree, and buckpucks are the most popular choice, but im still slightly confused on how many of each i will need for my particular setup,

i figure i may as well make DIY LED building my new hobby as my car is done and i quit world of warcrack

ok just a quick question...

i measure the voltage across my led's and it measures 3.41VDC and when it gets hot it drops to 3.39VDC.. but when i put a 1Ohm resistor is series and measure voltage across that it read 1.355 ??? how does that work?? isnt the max current of 1amp at 3.7V.. how does this work out, how can i be overdriving current but under driving voltage.. the same thing happens with both of my arrays.

my other ne is driven by a ELN-60-48 and Led's read about 3.5 but the resistor reads 1.5##V

so does it mean im running the leds at 3.4VDC @ 1.3## amps...

sorry what i meant to say is that if I wanted to to mount my leds to a 1/2in thick piece of metal, do you think I could get away without machining heatsinks? I have about 3/4in of materials on top of the heatsink but after tha fans, the fixture might be a bit thicker than I want which is a bit uncomfortable.

Those voltages at a given current are maximums that Cree sees during testing of batches. It doesn't mean that every LED has to run at that voltage. They are to be used as guidelines for designing arrays, not defacto law.

If you haven't already, turn your current down.



You could, but you will be relying on forced air cooling to keep the temps in check. I'm sending you a PM.

So has anyone done "spotlights" yet? What I mean by that is that on my 75g I run a 4 bulb T5 setup that isn't good enough for clams, so I was thinking about a small 2-3 led array to highlight the clam.

Thoughts?

Like this?

http://www.nano-reef.com/forums/index.php?...t&p=2443686

That or something similar to your LED actinic array.

do these bulbs put out as much light as the nanocustoms one for $120?

http://www.divinelighting.com/par-38-led-t...x-1w-p-275.html

oh and i found this in my search... what do you guys think?

http://www.ledwholesalers.com/store/index....at&catId=65

More LEDs won't make up for less power per LED, especially in the case for the 12W PAR38. As far as the other link, which setup were you looking at?

Do you think the XP-G lamps have any use for us, or would they be overkill like the MC-E? They claim to put out a good bit more light than the XR-E and the cost is about the same as an XR-E (ie, more bang for the buck).


http://www.cutter.com.au/proddetail.php?prod=cut937

The grow lights, I would think if it's enough light for plants to grow it should be the same for coral. The led fixture also seems to be pretty good too

Depends on the application really. At 1A, it's a fairly hefty jump in output. We typically turn the LEDs down anyway. One compelling reason to use them is that because of the greater efficiency, you can run them at lower drive currents for the same lumen output, and get less power consumption and less heat. The XP-Gs would be good on deeper tanks without optics, but there isn't a royal blue LED to match the output.


Plants and corals have different requirements. 1W LEDs are still 1W LEDs. It's still not quite enough for most tanks except the most shallow. For the cost of some of those lamps, you could DIY a real high power setup that will smoke them. The only one that is worth a damn is this one

http://www.ledwholesalers.com/store/index....p;productId=407