Come in and join me....

[neanderthalman]

LEDs are just light emitters. but the white LEDs are filled with phosphors (to make them white...) and such they emit rays that well emit heat. thats why i feel them at 1 inch away. but after that its basically visible light.

 

Uh....not quite.

 

Look up "blackbody emissions" and "planck's law" and you'll understand why there is still some radiant heat from LEDs. There will be radiant heat from the blue LEDs as well as the white ones, and it's not due to phosphors. The phosphors will cause the LED to be slightly hotter due to the inefficiencies associated with the phosphorescence. The actual emission of the IR, however has nothing to do with the phosphors, only the temperature of the LED. Lower the temperature, lower the IR emissions.

[maverick]

Any electrochemical reaction will produce heat. some just more than others. it is just the nature of the beast

[evilc66]

Hope fully this will now make what I'm saying a little more clear. Take a look at the picture below.

 

 

Link for a bigger version

 

This is a direct comparison of both LEDs. If you want to make it easier to read, assume that both LEDs produce 100lm. Remember, lumen measurements are taken directly over the center of the LED and is a measurement of point light, not total light that is produced by the LED.

 

I didn't take into account reflection or refraction because under 45deg off center, the losses are minimal. I also didn't bother to show the Rebel up to it's 160deg range. All this information is taken from the respective data sheets. This is without optics, mainly because people have claimed that the XR-E won't need them for this kind of application.

 

What you will notice is that the Rebel, even though it has a wider viewing angle can still put the same, if not more light INTO the tank. Yes, there will be light lost and bounced from the glass, but the same is true with the XR-E. But the edge of the tank is only part of the lighting scheme. Once you get closer to the center of the tank, wouldn't you want a wider light spread? A wider light spread will mean you can use less LEDs to cover a certain area. Less LEDs means less heat, and less cost overall.

 

Up to 20deg off center, both LEDs are pretty similar in output. Once you get past 30deg, the XR-E really suffers. 30deg is still useable light, with little to no light lost due to reflection. The Rebel has higher intensity at that angle than the XR-E. The situation gets worse past that for the XR-E. At 50deg (which is still useable light) the Rebel still holds about 60% of its intensity and the XR-E drops off hard. Granted, at 50deg and close to the edge of the tank, you will lose some light out of the glass, but a lot of light will be reflected back in. No matter how much light is lost through the glass, it's still more light than the XR-E.

 

Does this make a little more sense now?

[hyewiz]

you guys are amazing, i wish i knew half of what you guys know, i found this images for led project its amazing what some people can do

 

[Gomer]

Evil, you can't think of it in percents. If you can the Rebel at 350mA, you'd get the EXACT same percent image. you really need to think of integrated overlap intensities.

 

Chew over this one

 

Lets make some very crude number calcs.

Assume no refraction, no relfection.

A biocube has ~ 13" wide footprint. with the lights about 13" from the substrate. This makes a 1/2 angle of around 30deg. Below is an intensity chart comparing several LEDs. The absolute numbers are off a bit, but the relative numbers are good. Take a look at the intensity difference from 0 to 30deg.

 

[ccjung]

Sorry to show up late to the party guys. I've been following your work(s) and I can say that you guys definately know your stuff.

 

I havent read the whole thread, but if this debate is about the cree vs. rebels. I see both of your points.

 

Gomer - your statement about integrated light delivered is valid. Looking at the integral of the curves, I would say that the Rebel delivers more total light, although the Cree emits more at the center. However, I have to disagree with the importance of the light delivered at the wider angles, because most of that light (snells law) would end up scattering. For a straight head on comparison between emitters without respect to optics, the Cree wins.

 

Evil - I would say that the Rebels in an aquarium application would deliver more useful output to the tank if the proper optics are used to deliver the light in a beam that would minimize refraction.

 

One other thing to consider is that these measurements are in Lux, which are more commonly used for visible light. It would be interesting to see if they have any par measurements of a comparison between the 2.

 

Keep it going guys!!

 

Chris

[The Propagator]

you guys are amazing, i wish i knew half of what you guys know, i found this images for led project its amazing what some people can do

 

 

 

NOW THAT IS SLICK !!!!

 

Post the link to that website !

 

[evilc66]

Gomer, you are right about integrating the values under the luminous radiosity curve. That was something that I only really started thinking about this morning. And if you are to compare the top end Cree to the top end Rebel, yeah, the XR-E is going to win. I was trying to go down a more apples-to-apples aproach and assume that the two LEDs had identical output.

 

If Luxeon came out tomorrow and said "Hey, we have an LED that can match the XR-E!", would you not agree that the Rebel would actually be a better LED (in terms of overall light intensity), even without optics?

 

To get the absolute most out of the Rebel, you will need optics, no question. But, the Rebel as a stand-alone unit is still very much capable of putting a ton of light into the tank. Your losses really don't start getting significant until after 50deg off center, where the Rebel can still put light into the tank.

 

But sadly apples aren't apple in this case, and the XR-E overall is a better choice without optics. But that little ole' Rebel ain't going down without a fight.

 

 

 

 

 

 

 

Don't know if you know this, but tightening the viewing angle is a trick LED manufacturers have been using for a while to bump the output of T1 and T1-3/4 LEDs. Wide angle, high brightness LEDs are not easy to make, not saying that high power ones aren't easy either.

[Gomer]

Actually, it is more apples to apples if you consider that the total lumen output (integrated over a sphere) are about the same. It is the distribution that changes.

 

Without optics, I would still go with the Cree, Luxeon is rather flaky with product launches, promises and consistancy compared to Cree. However, Luxeons tend to be cheaper.

[evilc66]

In the same token, the Luxeons are better on color degredation and longevity than most other manufacturers.

[coolwaters]

to me luxeon are durable and can withstand being overdriven better then other brands. but at the same time their sacrificing efficiency. cuz when K2s came out which was at 40lm per watt cree and other brands like Edison are already at 60lm+ but when rebel came out things changed. and the new K2s.

 

btw that pic above with the 100+ leds are they flux LEDs? they dont look like the 3mm, 5mm, 10mm leds

[SaltyJohn]

Ouf!... I leave you guys alone for a few hrs/days on earth and I come back to find you on the moon!....

 

Great info on both sides, it's been an education for all of us I assume.

 

Now my input, far less science into the matter, just more hands on experience.

 

If I had two LEDs (XR-E/Rebel) that put out the same lumens but one at a wider angle than the other, I would always buy the wider angle one, since I can improve it a lot more with optics that I can improve the narrow angle one, since -as Evil mentioned- manufacturers, do tighten the angle to boost nominal lumen values.

So point (1) if all else is the same, buy the wider angle

 

On the other hand, there is so much more to look at.

Cost (someone mentioned one of the LEDs being cheaper to buy) here is a total of:

 

1) Actual LED cost

2) Current chosen to operate at (since we are consuming electrical power, Cree @ 350mA is a bargain)

3) Optics ( I can't say I have seen many choices for the Rebels, and Phillips doesn't mention any approved lens yet. For the Cree's I can get descent lens for under a buck)

4) Cooling, where junction temperature and the LEDs efficiency as Tj rises is of essence in choosing proper heat sinks. (Rebels keep more than 90% brightens at 67C degrees, XR-E at 65C -no biggie- but Seoul's P4 at 72C. So the P4 needs less efficient cooling).

5) Life span of LED in direct relation with (2) working current rating.

 

So.. a bit more complicated that upfront purchase price.

 

The good part is that we are discussing a very cutting edge product, with blinding fast development phases, hence while the Cree Q5s are as good as I need, the release of the -way too expensive, I agree- R2 bin will drastically drop the prices for Q5s.

 

An isolated example, but its like buying a PC.

I paid for my first Sony Vaio Pentium (Pro) MMX 200MHz around 4.000$, in.... 1995/6?

You all know what you can buy with 4.000$ today, especially if you consider that 4.000$ then was equivalent to roughly 9.000$ today's values!

 

So, point (2) do not buy the latest and greatest. It is too expensive and it wont be the greatest for more than a couple of weeks.

 

All in all, I think the Cree today (and I can't stress 'today' enough) is the best bet for my money.

 

Another important issue since we are using these in our reef tanks is color temperature. Slowly but steadily, we are seeing LEDs with bins at temperatures in excess of 6.300 Kelvin, but still we are a loooooong way from throwing our Actinics out.

[Gomer]

oh, and I would like to add that you need to consider working with these things. Soldering can be a PITA (the cree's do not want to solder up on the top pads for me). If you are using emitters on little boards, keep in mind that there have been studies showing that they (at least the ones tested) are not that efficient and cause local heating problems.

[evilc66]

I agree Salty. We have been focusing on only one specific parameter that should be weighed in when choosing an LED. Sadly, all these options make choosing one very difficult for the uneducated. But some of these variables can easily be skipped over. Heatsinks nowadays are plentiful and cheap, and even the hottest K2 can be cooled with ease.

 

While power is an issue, it's not really the deal breaker it once was. There is virtually no cost difference in the drivers, and changing from 350mA to 700mA drive current isn't going to hit your electric bill in any significant way unless you are running a very large array. For example, my LED setup that I will be putting over my pico tank will consume 29W of power @ 700mA. Based on my recent power costs, it will cost me $0.51 to run it for 10 hours a day for 30 days. At 350mA it would cost me $0.26. Is anyone really going to worry about that? Pumps and heaters will consume more than that. Granted, bigger setups will consume more power, but even to get the same light as a 70W MH, it will cost me a little over $1 a month.

 

I'm not trying to say that these are not possible factors in choosing an LED, it just starts making a hard decision even more confusing.

[coolwaters]

if your guys are trying to find the break even point in LEDs to MH....dont...its the cost of changing bulbs that break even within 5 years.

 

thats like 6 MH bulbs

and your LEDs will drop to 50-80% output.

[coolwaters]

evilc66 hows your LED project coming along? im done with one of mine. i'll post a pic of it in action.

[SaltyJohn]

Evil and coolwaters things are not exactly like that... Granted for nano/pico setups these costs are easily ignored next to the cost of even 4-5 hermit crabs or an urchin - and yes this is a NanoReef forum, so forgive me if I am drifting to the wrong direction.

 

Power consumption is the main reason that pushed me initially to look at alternatives to the beautiful light that MH gives out -with great glitter I might add.

 

Driving the LEDs at 350mA or at 1000mA is indeed trivial if you only look at the drivers and the circuitry in a small tank. But a 250W array at 350mA (~230 LEDs) becomes a +700W array in a medium tank. 450 extra Watts are not pennies worth of ConEd bill..

 

Another truly important reason is where the heat generated, lies. With MH it is all over my water, negating a heavy chiller in the summer, with LEDs it is only in the room's air (exhausted out with minor ducting), which takes out of the picture my 800W chiller, temp controllers, the extra fans, and most importantly the extra evaporation.

Hard to put $ signs on those, but I know it has been a pain for me to balance all that extra heat...

 

With bulbs (both MH and LED) prices dropping, and with electricity cost of KWh going up, it is not only the fewer bulb changes that one will benefit from here!

 

Anyway... not to babble on, point is that LEDs bring more to the hobby than 'sexy technology'

[evilc66]

Salty, I did mention in my last post that large arrays are going to bump costs up quite a bit. But like you so rightly mentioned, once you factor in the lack of heat generated by the LEDs that will negate the use of a chiller, things can get into a cost savings. It's going to be a long time before LEDs can truely be a cost savings in comparison to any other light source.

 

One thing to take into consideration that we really haven't talked about much, is that while we are shooting to match brightness levels with MH and T5 setups, we really don't have to. PAR levels from LEDs are greater than that of MH, and as a result we can use less light to get the same PAR level. While the brighness is great for viewing purposes, it's not required for growth.

 

 

 

Coolwaters, I have actually been doing quite a bit of work on it since school is out for the next couple of weeks. The frame for the hood is almost done, and I am going to be rewiring it later today. I'll try and get some pics up later today. Maybe I can finally post some of my stuff on my own thread for once

 

Only kidding guys. This have been a great discussion on LEDs that will benefit everyone. I don't think I have seen a discussion on LEDs go this in depth on any other forum before. Keep it coming.

[coolwaters]

yeah cant w8 to rap this up with a final LED post...

bottom line that im trying to give is that anyone can get a high power LED array for a nano tank under $100.

[evilc66]

Time for an update!

 

Finally got some time over the weekend to finish things up a little. I still have a little to do, but it won't take long and should have it done by the end of the week.

 

 

So here is a shot of the tank lit with the new fixture. You can see the frame on top of the tank.

 

 

Slightly lower shot showing my sole inhabitant of the tank so far I got it out of one of those fish bubble lights to see how the flow was working out. He's a little heavy, but moves around pretty well. I think the flow is borderline, and may switch out to the next size up pump.

 

Please excuse the cloudy water. I have been leak testing the tank and sump together for the last few days with tap water.

 

 

This is the enclosure I made for the regulator circuit. I have a small 40mm fan pulling air through the box keeping everything cool. The FETs are clamped to the lid as a heatsink and runs ice cold. Temps don't ever get above room temperature. I'm really pleased with how that turned out. The fan is a little noisy though, and I may drop its voltage a little to keep it quiet.

 

 

Here is a top shot of the frame. It's all made from 3/4"x1/8" aluminum angle and fastened with stainless hardware. It might be a little overkill (much like the heatsink), but it's rigid and light weight. The heatsink weighs more than the frame does.

 

The frame is not quite done yet. I still have to clad the outside of it with acrylic to match everything else.

 

 

This is where all the magic happens. You can see the surface of the heatsink where I have space to upgrade my LEDs. I can get more light than a 70W MH if I fully populate everything. I will have a peice of acrylic covering everything.

 

Airflow (more as a just in case) will be a low profile 80mm fan that will also circulate air around the LEDs and the heatsink.

 

 

Overall, I am really pleased with how this is turning out. The next challenge is going to be with the tank itself and getting the aquascaping set up so it will hide the plumbing.

[coolwaters]

lol i thought your just doing 4 LEDs but i just realized that each board has 3...

 

nice set up btw hope to see some coral soon.

[evilc66]

I'll be getting sand and LR on thursday. It will be a little while before I can get coral in, but I have some stuff groing in my BC14 that I will probably frag and put into the new one. I'm excited.

[coolwaters]

ic. im getting all my stuff hopefully btw next week...

my parents arent going to be too happy about this =P

[SaltyJohn]

Evil.... I really like your setup!

 

Heatsinking to spare, lots of room for LED expansion, enclosed controllers/electronics with own cooling... I like!

 

I might have missed it, but I would definitely put a splash guard piece of acrylic between those LEDs and the water!

 

Can you elaborate a bit on the 'regulator unit enclosure'?

[coolwaters]

evil where did u get that heatsink? i want something that big.