DIY LEDs - The write-up

Status
Not open for further replies.
It seems that the CREES seem to be the ones everyone is buying, are they the best lights or just the most cost effective for the price?

What about the Endor star or the ZxPower Z50 50W High Power Led - White, 30-36v, 1.5A, 3000lm [ZxP50] (May be a bit much at 3000lm)

I am just trying to understand if there is a reason for the CREE trend
 
Just got off the phone with someone at the local radioshack, he said they carry thermal heatsink compound for gluing heatsinks to CPU's. Does this sound like it would work?

I know my local radio shack sells "arctic silver" thermal compound which can only be used if you screw down the star. It does NOT have adhesive properties. They might also sell "arctic silver" thermal adhesive which has adhesive properties and can be used to "glue" them down.

Here is where they can be purchased or viewed so you can see the difference:
Arctic Silver Thermal Adhesive
Arctic Silver Thermal Compound
 
It seems that the CREES seem to be the ones everyone is buying, are they the best lights or just the most cost effective for the price?

What about the Endor star or the ZxPower Z50 50W High Power Led - White, 30-36v, 1.5A, 3000lm [ZxP50] (May be a bit much at 3000lm)

I am just trying to understand if there is a reason for the CREE trend

The Cree trend is probably partially habit, and partially justified. Arguably, Seol Semiconductor and Luxeon also produce suitable LEDs. It probably depends on your criteria. IME the following are important features:

1) Efficiency - lumens per watt. Cree XR-E, MC-E, XP-E, etc and Luxeon Rebels are all around 80 - 90 lumens/watt the way we drive them. Cree XP-G are around 120 - 130. Some people used to use the older Luxeon I/III/V series, but they're far less efficient. SSC make some LEDs that have similar specs, but they fail the other two criteria:

2) Usability in our standard approach: Some SSC products, and even some Crees, fail here. Generally, there are a whole class of super-high-output LEDs that are just multiple emitters on the same die. Some of these, like the Cree MC-E, are easy to use, because the emitters are in series. Some, like the SSC product that's escaping my mind right now, are in parallel, which means they need drive currents WAY above what our typical drivers can handle. Also, some of these super HP LEDs just put out a TON of light from a single point source, which makes positioning/heatsinking/spreading light out over the tank difficult. And generally, they're no more efficient than using several XR-E for the same output, so there's no argument to use them.

3) Price/availability. SSC make some LEDs that are similar to the XR-E in nearly every spec, but they're generally very hard to get at reasonable prices through the typical channels we're using as hobby enthusiasts, which probably accounts for their lack of popularity.

You mention the Endor Star. Besides the Cree products, I find it to be the most viable. It's just Luxeon Rebels pre-mounted on stars. I've used it in several builds. Though, it does bring up some more selection criteria beyond the above - the Rebels have a wider viewing angle than XR-E, so the light will be more spread out, and optic choices are a little more limited. The current king of HP LEDs, the Cree XP-G, "suffers" from these two issues as well - it has a pretty big viewing angle, and no wide optics to "fix" it.
 
stealle; Thanks for that UV LED info I was un-aware.


Wid; Often when you don't have a high enough power iron you end up with a flux mess. A sort of runny pool that prevents any actual soldering to occur. Then you go back and try again and again and it gets worse and worse.

At this point I think the suggestions to abraid the surface is what you need to try. Use sand paper or a jeweler's file to "reach the shine". Then pre-solder like Stu likes to do and then come back with a pre-tinned wire.

What wattage is your iron and what type tip is it?
What solder are you using?

I have sort of already tried what you are referring to, but to no avail. Maybe I haven't been going hard core enough at it? I took an old kitchen knife (got a set of new wusthoffs for xmas this year :uzi:) and scraped it shiny a few times but that didn't do it.

My iron is a butane hot-as-the-sun iron, made fast work of the dealextreme and nanotuners stars. And my solder is rosin-core but that's all I know about it. I've had it for years...

I'm almost at the point where I feel it would be easier to somehow epoxy the wire down onto the contact for these five stars... :lolspin:
 
My iron is a butane hot-as-the-sun iron

IME the iron's temperature control and stability are more important than the absolute temp it can hit. If it hits 300c or whatever but doesn't have the mass in the tip or wattage to hold it when it's touched against the star, it doesn't matter.

And I am serious about wondering if the solder they plate those stars with is lead free, since it acts like it (wants really high temps, refuses to stick, etc.)
 
It seems that the CREES seem to be the ones everyone is buying, are they the best lights or just the most cost effective for the price?

What about the Endor star or the ZxPower Z50 50W High Power Led - White, 30-36v, 1.5A, 3000lm [ZxP50] (May be a bit much at 3000lm)

I am just trying to understand if there is a reason for the CREE trend

That's 60 lumens per watt. Not very impressive efficiency. The 3 watters from Cree best that easily. If efficiency doesn't matter and only lumens matters, then by all means get a couple of these if they work in your setup. You're also unlikely to get optics for that ...
 
Der_wille or any one else that has used LEDs. Can you give me an idea how many LEDs you would use for 3 of my tanks. I have 4 39watt T5 above 2 of the tanks now. The third tank is the refugium/sump with a lights of America Compact fluorescent bulb on about half of it. All three measure 45x23" with 11" water depth and are stacked on top of each other.
I am keeping mostly anemones and SPS in one and I have 9" above the water total for clearance. The second one I have anemones and softies and about 14" above the water .
The refugium has about 14" above the water also. I would like to keep the light in the tanks and off the glass as much as possible. I prefer not to use optics if I do not have to. I was thinking I could machine a slight angle into the head sink or use aluminum channel at an angle for the outer LEDs to help keep the light in the tank.

I want to get in of the group buy at nano-reef while the LEDs and Meanwells are cheap.
Would 26 LEDs each be enough light for the top two tanks?
13 XP-G CW
13 XP-E or CREE XR-E RB (which is better?)
2 ELN-60-48-D

How about 13 XP-G CW for the refugium? Is there a better choice for here?
 
laverda,
if you want to control the light output such that you can keep light off the glass, optics is the best way to go. Without secondary optics, these LED output angles are pretty wide and I can almost say for sure you will get light on the glass no matter how you angle them...
I'll let someone else answer your last question, since "how much is enough" is a very difficult question to answer... :)
-R
 
That's 60 lumens per watt. Not very impressive efficiency. The 3 watters from Cree best that easily. If efficiency doesn't matter and only lumens matters, then by all means get a couple of these if they work in your setup. You're also unlikely to get optics for that ...

I was joking about the ZxPower Z50 50W, but did want to understand the desire for the CREEs vs the endor star. Just to understand the issues and the cost.
 
I was joking about the ZxPower Z50 50W, but did want to understand the desire for the CREEs vs the endor star. Just to understand the issues and the cost.

In terms of XR-E vs Endor Star, efficiency and other specs are about the same, but viewing angle is wider on Endors. XP-E is a little more similar to the Endors. The XP-G is more efficient than all of these, by about 30%, but also has a fairly wide viewing angle.

Price is close between Endors and XR-E.
 
Santoki
I should have mentioned I currently have my T5 about 2-3 inches above the water and can do the same with the LEDs, if need be. It would be easier to work on the tank and view from above if the lights were higher. What optics do you think would work best for my aplication? It looks like only 40 degree are available for the XP-Gs.

laverda,
if you want to control the light output such that you can keep light off the glass, optics is the best way to go. Without secondary optics, these LED output angles are pretty wide and I can almost say for sure you will get light on the glass no matter how you angle them...
I'll let someone else answer your last question, since "how much is enough" is a very difficult question to answer... :)
-R
 
So i've looked through many pages of this thread to figure out a trend in how many LED's do you need for a certain size tank. I believe it was der willie zur macht who had a post many pages ago that gave a rough estimate of 1 LED for every 10 - 20 sq. inches. that tends to be what i get when i calculate what most people say they are going to do. most have been in the 12 - 15 sq inch range it seems.

So my question is for a 125gal 72x18x23 that I will be starting in a few months. From what others indicate they are planning on their tanks I would then need anywhere from 65 - 130 LED's. Ok, for starters I think I can afford to build a fixture with 72 LED's. That would space them side/side roughly 2.5 inches at 24 per row giving me 3 rows. But the 6 inch average between rows seems a little large compared to what others have done. Now I go back to reminding myself this will be a fresh build, won't get to the harder corals right off the bat so maybe i can start with 72 spaced as mentioned and then 6 - 8 months after i get going, add another couple of rows. For example if I don't go the LED route i'm pretty sure I'm going to go T5 with a two bulb fixture the length of the tank. Or two 3 ft fixtures. Anyway, gives me two bulb rows. Then as I need, add more T5 fixtures, but to start, the tank is lit, the fish can see and some of the easy corals will probably be just fine. My goal here is to always just use LED's over the tank, from day one, even if I start with a half load of them and add -on as I can afford it.

Would the 72 LED's, maybe up 8 inches from the water line, give me enough light and enough spread to adequetly light the tank so I can start off with LED's from the get-go?
 
schatzi I think your estimations are pretty acurate. Better then mine where, I need about twice what I was thinking. 2 T5s would defianatly not be enough light for a tank your size.
 
They won't really be spaced 6 inches between rows. that is if your lights are at water level. Since your led's are higher it won't be that way. For instance, I have a 120G tank (60x18x26), I am using 110 LED's on 2 heatsinks (24"x8.5"). I have spacing around 1.5 in. between rows.

Assuming you use a 5 foot heatsink, it really doesn't have to span the entire length (8.5" wide), with 3 rows of 24 you should have spacing between rows of 2.5". If you plan what type of optics you want to use and how to orient the light beforehand you might even be able to get away with a smaller footprint.

IMO it's probably better to get it all over with in one shot than to do 72 now and then a few more rows later. It will be a PITA to take it down, redrill, more soldering, more wiring, rearranging driver wiring etc. etc.

Up front cost is a tough pill to swallow but tis better to cut once and cut deep than to do it a bunch of times.
 
g8gxp brings up a good point - when thinking of how many LEDs you will use on your tank, don't get hung up with spacing between the LEDs. Determine how many you need based on how much light you need, then just space them evenly on heatsinks that come within a few inches of the tank's dimensions.

Many people are quoting installed LED spacing (2", 2.5", etc) on their heatsink, which DOES NOT translate into LED density per unit of tank area - because, typically, the heatsink only covers a portion of the tank.

Doing the whole thing now vs. doing some now and some later would, for me, depend on a few factors:

1) Overall design - if you do the typical one or two monolithic heatsinks, then it would be a little tougher to add light later. If you do something more modular, it would be considerably easier - say, 5 or 6 12"x6" heatsinks instead of two 24"x10" heatsinks, for instance.

2) Time and money. If it'll be more than a year or two before you need the additional light, might as well wait. That's a long time in LED development terms, and prices may have come down enough by then that the additional LEDs are much cheaper.
 
Status
Not open for further replies.
Back
Top