*DIY* High Power LEDs

[neanderthalman]

Sorry Coolwaters, I should have explained it better. The angles in optics aren't measured from the surface, they are measured from a line normal or perpendicular to the surface. Here's a diagram that should help you see where the angles are.

 

 

BTW, this is for a school project, right? Might be worth discussing reflection from the surface in your presentation . Just steal my graph and use it to show how reflection is minimal.

 

2.4% of a very bright light source will still look bright to the human eye. The human eye is terrible at gauging the intensity of a single light source, as has already been mentioned in this thread.

 

Edit - thought I'd add. theta-i in the equations is theta-1 in the diagram. theta-t=theta2.

[brshriver]

Sounds like we need a PAR meter. I have one, but doubt that I am close. Apogee's runs about $300.

 

PAR at the ocean's surface at mid day at the equator is around 2000 I think. Here in MD, I see 1700 to 1800 on the one day that I checked. In my tanks, numbers are MUCH lower than that even under 250W MH lights close to the surface. In the 200 to 400 range. I still need to plot out the light gradients over my whole tank though.

[Coyote]

Hello,

 

Here is a small simulation of 4 K2 (2 white and 2 royal blue) over a 30cm wide tank.

I take in account reflection into salt water and glass. there is no lens or reflector over the leds

 

 

Only the ray that represent more than 5% of the incident light are represented. as you can see only a small amount of light goes out the tank.

 

I'v also made a small test over my k2 spot. A luxmeter placed just outside the tank near the bottom of the glass read less than 20lux. The same configuration without the tank read more than 5000lux

[HecticDialectics]

I'v also made a small test over my k2 spot. A luxmeter placed just outside the tank near the bottom of the glass read less than 20lux. The same configuration without the tank read more than 5000lux

 

 

Can you use your meter underwater? That would be very interesting.

 

The total angle for the K2s is something between 160 and 170 degrees for 90% of the light if I read the spec sheet right. You'd want the leds sitting practically on of the water without lenses, but that also ignores how much light intensity is lost as it travels through the water back and forth. The greater the angle the further it has to go before it hits anything. For some reason, aquarium depth seems to affect light intensity much more than comparable numbers I've seen for transmittance for ocean depth.

 

I just played around with a flashlight focused into a tight beam, and it's great for getting an idea of what's going on. I wish I had a laser pointer haha.

[coolwaters]

neanderthalman since i have a physics class understanding anything that is related will help for sure.

 

about comparing the sun to any of our lights gives pretty depressing results...

all we can do is get enough to keep reefs alive without killing our wallets.

 

Coyote did u design that? must have taken a while.. but i like how it explains where the light goes.

 

btw is that 160 degrees that the K2 are giving out?

and would a 25 degree angle be any better?

[HecticDialectics]

It looks to me like those are only 60 degree angles off center (if you look closely at the dot matrix it's just about a 3-4-5 triangle) thing in coyote's diagram, which would make them 120 degree spreads.

[Coyote]

Don't worry, i've used an optic software to draw this simulation... to much calculation to do it by hands...

 

If you check the datasheet of the k2, you can calculate that 75% of the light output is inside an angle of 45° half angle (90% inside 60°).

 

I also tried a lot of different lense for k2:

-l2optics 3°

-l2optics 6°

-l2optics 10°

-fraen 15°

-fraen 25°

 

None of this are really useable. finally the best is to use a 45° lens or a reflector.

 

Otherthing, i've just received today some cree XR-E white led. waouh, they give a lot more light than k2. I will surely try them on one of my spot soon....

 

hecticdialec: yes i've only drawn the ray inside 60° as it's no use to draw the other...this then represent 90% of the total output...

[coolwaters]

so a 45° would be best or a reflector?

 

i was thinking about the reflector since its cheaper and i dont want my light to get too patchy. thanks i think u just saved me a lot of money XD.

 

btw where did u get the cree LEDs? and how much?

[Coyote]

so a 45° would be best or a reflector?

 

i was thinking about the reflector since its cheaper and i dont want my light to get too patchy. thanks i think u just saved me a lot of money XD.

 

btw where did u get the cree LEDs? and how much?

 

I've made the calculation and for a final angle of 45°, reflector and lenses have the same efficacity... it's around 85%

 

If you want a final angle above 45°, a reflector is best...

under 45°, go for a lense.

 

I've also tried a new concept. If you use k2 with l2optics lens 3° you can light a 30x30x30cm cube (27l=~8gallons) from the ceiling!!!

 

As for the cree, i've had them at 7$ from a french reseller. It's pretty expensive but i only bought 3... Finally i think it's interressent as the light output is really better and you do not need any lense or reflector...

 

If i was to make a new spot i would go for a mix of white XR-E and royal blue K2

[coolwaters]

dang for $7 i can get 2 of the best cool white K2's. so i'll probably stick with thos.

 

i just wished that the rebel was like 100 times bigger so actually hold it with my finger tips...and it sux cuz the power input is on the bottom along with the thermal pad... who designed that???!

 

does anyone know where to get some cheap LED drivers at a good price? PC PSU just isnt my thing...and i would have to get regulators which cost even more.

 

this project now is pretty much money and finding the right things...

 

Coyote did u try to use thos 3 degree beams and aimed them at the same spot to see how close u are to the suns rays?

[Coyote]

one of the advantage of the XRE is that it doesn't need any lens so you have to take it into account.

 

I did try the 3° beam lenses, they are very impressive but i didn't try to compare them with the sun...

[strangelove]

Sounds like we need a PAR meter. I have one, but doubt that I am close. Apogee's runs about $300.

 

PAR at the ocean's surface at mid day at the equator is around 2000 I think. Here in MD, I see 1700 to 1800 on the one day that I checked. In my tanks, numbers are MUCH lower than that even under 250W MH lights close to the surface. In the 200 to 400 range. I still need to plot out the light gradients over my whole tank though.

 

 

Right but that's par on the oceans surface, most reefs are 5' to 10 feet below the surface of the water. What are par readings 5' to 10 feet below the water surface. It will be much lower, maybe our resident caveman might know the answer to this one.

[neanderthalman]

Sorry, not off hand, and I've gotta get to work. However, ocean water over a reef is typically pretty clear, so absorption and scattering are fairly low. The reflection, as already discussed, is minimal, so at 5' to 10' below the surface, you'll find that the sunlight is not significantly attenuated.

 

Now, reefs extend down to about 30 meters (~90') below the surface. After depths like that, there is plenty of absorption/scattering and light levels have dropped significantly.

 

I'll look into it tonight and see if I can give you a better answer than an estimate. It's kinda complicated due to the wavelength dependence on absorption.

[HecticDialectics]

Check out the bottom of the article...

 

http://advancedaquarist.com/issues/august2003/feature.htm

[coolwaters]

he need more LEDs lol or else he'll have to over work them at a high voltage.

my AC transformer can only take about 4 bulbs (15w about) without totally melting. so i'll have to add another power supply or get a better one.

 

its pretty hard to find power supplys. if its output is too low its usless and ifs its too high the current isnt constant...

[neanderthalman]

nice find hectics!

 

That data is pretty much what I was thinking. I thought the absorption in seawater would be a little less than they reported, but it's roughly what I expected. At around 1/3 of 1m in depth for a typical nano, the losses to absorption will be in the region of a few percent. That's really not significant in the grand scheme of things.

 

The light loss due to depth in nanos is due mainly to the distance from a local point source of light. Light intensity diminishes proportionally to R², where R is the distance. In the case of a light mounted 4" from the surface of a 12" tall (10g) tank, the bottom of the tank is four times further from the light than the top of the tank. Because of the R² relationship, the light will be sixteen times more intense at the top of the tank than at the bottom of the tank (without a reflector or reflection from tank walls).

 

Compare that to the sun, which is a distant point source. Under natural sunlight, the surface of the ocean is 93 million miles from the light source. The reefs, ten feet below the surface, are 93 million miles + 10 feet. The difference between the two is insignificant, and so is the loss from the R² term. The primary loss of intensity in the ocean is from absorption, not distance.

 

So, comparing the rate of light loss with depth in our tanks to natural conditions isn't really appropriate, since the mechanism of light loss is completely different in the two circumstances. It's good to have some oceanographic data to show that absorption/scattering is fairly low, so we can focus on what is really happening in our tanks. It can't be directly applied, however, due to the close proximity of the light source.

 

Now, how does this relate to LED's? The tighter the beam angle, the more closely the LEDs approximate a distant point source. This means there is less loss of intensity from distance. Since absorption, scattering, and reflection are all minimal, tight angle LEDs will be better at penetrating deep tanks than other lighting sources.

 

Compared with conventional fixtures, even at a lower intensity at the water's surface, an LED fixture may have a higher intensity at the bottom of the tank because of less dependence on the R² term.

[strangelove]

nice find hectics!

 

That data is pretty much what I was thinking. I thought the absorption in seawater would be a little less than they reported, but it's roughly what I expected. At around 1/3 of 1m in depth for a typical nano, the losses to absorption will be in the region of a few percent. That's really not significant in the grand scheme of things.

 

The light loss due to depth in nanos is due mainly to the distance from a local point source of light. Light intensity diminishes proportionally to R², where R is the distance. In the case of a light mounted 4" from the surface of a 12" tall (10g) tank, the bottom of the tank is four times further from the light than the top of the tank. Because of the R² relationship, the light will be sixteen times more intense at the top of the tank than at the bottom of the tank (without a reflector or reflection from tank walls).

 

Compare that to the sun, which is a distant point source. Under natural sunlight, the surface of the ocean is 93 million miles from the light source. The reefs, ten feet below the surface, are 93 million miles + 10 feet. The difference between the two is insignificant, and so is the loss from the R² term. The primary loss of intensity in the ocean is from absorption, not distance.

 

So, comparing the rate of light loss with depth in our tanks to natural conditions isn't really appropriate, since the mechanism of light loss is completely different in the two circumstances. It's good to have some oceanographic data to show that absorption/scattering is fairly low, so we can focus on what is really happening in our tanks. It can't be directly applied, however, due to the close proximity of the light source.

 

Now, how does this relate to LED's? The tighter the beam angle, the more closely the LEDs approximate a distant point source. This means there is less loss of intensity from distance. Since absorption, scattering, and reflection are all minimal, tight angle LEDs will be better at penetrating deep tanks than other lighting sources.

 

Compared with conventional fixtures, even at a lower intensity at the water's surface, an LED fixture may have a higher intensity at the bottom of the tank because of less dependence on the R² term.

 

Wow I'm glad your here to explain that to us Great article too Hectic, peace. If I'm understanding you correctly, LED light is more focused allowing it to transmit through water further than diffused light from PC lights or scattered light like MH.

 

So I guess the main observation here is sunlight is drasticaly different than artificial light from, CF, PC, MH, and LED. Than the question should be, how do we draw a relationship between artificial light sources and natural daylight?

[coolwaters]

how do we draw a relationship between artificial light sources and natural daylight?

 

u compare the wavelength and try to get the artificial light as close as possible to the natural light.

 

but studies have shown that very short wavelength (blue) and long wavelength (red) has a greater growth effect then the other wavelengths. but since the blue looks better IMO people got into the actinics and blue moonlights.

 

even i tested that the blue light is better when it comes to growing things. 470nm is the peak i think...

 

 

------

 

in matter of the LED lights i found that my ac adaptor can only power about 20watts safely. i'll need to get a better adaptor or get 2 of them. which isnt sooo bad.

 

the array that i had before which was 4 LEDs at 3.75v at 1300mA which comes to 19.5 watts which was pretty close to the max...

 

now i have 6 LEDs (2 burnt out and 2 became too dim both my bad...)

 

right now theres 6 LEDs running at 4v each at 1300mA which comes down to 31.2 watts...with a 2 watt PC fan...O.O

no wonder my adaptor got soo hot after 20mins...

 

so right now i'll be testing the array with 6 LEDs running @ 3v on 1300mA 23.4 watts with a 2 watt PC fan.

which isnt too bad. if it gets too hot for my taste then im going back to 4 LEDs.

[coolwaters]

i bought a lot of heat sinks with fans on them....waste of money but watever fun investment....

its for computer processors so i can put quite a few LEDs on them. which is pretty cool (make it very dense) along with some class 2 transformers....i found some that are rated 30 watts+.

[coolwaters]

heres a picture update.

 

im comparing 4v to 3v (big difference) and other things

 

All pictures are 6 LEDs the only difference is the voltage.

 

4v compared to my 15 watt T8 bulb (the box says 20000k but it looks like 10000k or even 6500k)

 

father shot

 

4v compared to a 20w PC 3000k-4000k kelvin temp. (above)

 

4v close up (changing landscape soon....) added chaeto a while ago and theres lots of pods crawing around.

 

3v close up

 

 

conclusion....3v suxs....

i'll be running them at 3.7-4v.

the kelvin temp for the 4v is pretty high...its hitting purple...

sorry some of the pictures are blurry im going to get a new camera soon...

[lgreen]

what a nerd fest. lol. cool project though.

[coolwaters]

does anyone know how to reduce a 15v 1000mA 35 watt input class 2 AC adatper to 12v? is there a resister can is strong enough for that?

[Coyote]

I think it's really a bad idea to drive the led by voltage. The k2 (and xr-e, seoul...) have a lot of difference in the U/I response.

 

For example if you drive 2 leds at 4V, one can draw 1000mA when the other only 750mA. As only the current is important in the resulting light, you will not get the best of your led...

 

So use your 15V adapter and had a regulator plus a resistor in current mode. Any regulator will do:lm317, mic29152... For the resistor, choose any 1,25Ohm 2w you can find.

 

With this you can power 3 K2 in serie and you will sure they will draw exactly 1A.

[NanoClown]

lol lgreen ur right, but i wish i had the brains to do something this cool, nice job coolwaters

[coolwaters]

where do i get regulators? is there a good website that sells it?