DIY LED lighting MH replace

[xtrstangx]

Sounds awesome, but...

 

[rsxs1212]

taggin along

 

[Karl K]

tag

 

[CoolUsername]

Cool idea. Tagging along for the ride. Should we start-up a collection so speedstar can buy himself a digi-cam? JK.

 

[miatawnt2b]

along for the ride.

 

[FloridaFishMan]

I would like to see some pics too!

 

[Qwiv]

<a href=showthread.php?s=&postid=8060375#post8060375 target=_blank>Originally posted</a> by em21701
Just a note on driving LED's, they are best driven with a constant current. The constant current significantly reduces color and brightness differences due to forward voltage variation of each LED. If you want a dimming feature use Pulse Width Modulation (PWM) of the constant current. By varying the duty cycle of the PWM you'll dim the light accordingly with no risk of under/over driving the LED's. Qwiv, I would assume that the PIC you are using is doing just that? I'd love to know more about how you are using the PIC, does it have and RTC onboard or did you use and external one? Care to share the source code? I have been studying LED lightinig for a while, I hope to get into building a setup myself.

Basically. Look at the boost puck drivers as they have the dimming feature and you connect the controller on to it. OK, little bit more than that, but you get the idea.

<a href=showthread.php?s=&postid=8060433#post8060433 target=_blank>Originally posted</a> by speedstar
I bet them are real bright I thought about the K2's first and backed down to the 3's. Looks like you have already taken this the direction I was planning to with the pic chip. I was gonna see if they will work with SPS's first. By other blue do you mean the more purple royal-blue @455nm? Thanks

Actually, both the Royal Blue and the Blue do. You picked the cyan due to the Lux, but if you convert the Lux to PAR, by assuming all light is emitted at the spectrum peak (which is not very accurate) you will see the other blues produce more PAR and look much better. I tested this with a meter as well. LUX figures weight light in the 555 nm spectrum, so a light at 420 nm spectrum with = PAR to a light at 555nm will have a much lower LUX lever compared to the 555 nm light.

 

[exoticaquatix]

way cool guys, it was just a matter of time.
also, i finally know what the heck easttn's avatar is!! haha
-nick

 

[speedstar]

<a href=showthread.php?s=&postid=8069272#post8069272 target=_blank>Originally posted</a> by Qwiv
Basically. Look at the boost puck drivers as they have the dimming feature and you connect the controller on to it. OK, little bit more than that, but you get the idea.

But the amount of boost puck drivers needed for 150-LEDS gets crazy. Pics will be comming I have a camera just wish I could buy time. I tore it down this weekend as to properly affix the stars to heatsink and vent add fans. Should have complete pics over tank and off tank this weekend.

 

[Qwiv]

My application is smaller so this was the best route for me. With 150 LED's you will need to come up with something custom.

What is the foot print of your LED light with 150 leds? That must be pretty big, and a lot of $ in LED's!!

 

[dez_2121]

ok sounds like a lot of fun but i have to say you guys lost me a long time ago

 

[toastman]

I have been watching and waiting.....now where are the pix?:hmm1:

 

[pulse13]

Speedstar, does the $250 you paid for the LEDs and power supply include optics?

Qwiv, I'm interested in using a PIC contoller for an LED project I'm working on. I'll probably be using 2 Lamina Ceramics Titan RGBLEDs driven by a Luxdrive buck puck for each color channel. Can your PIC controller give input into the buck puck for dimming? The RGB Titans are 25w LEDs with multiple red, green and blue chips in each one and seven 60 degree reflectors built in.

They only have a press release on their site but they are taking orders and I've already received a RGB Atlas LED to play around with.

www.laminaceramics.com

 

[Qwiv]

Yep, you basically use the PIC to controll the input on the dimmer of the buck puck. You will need to play with teh values as it is not a linear value.

What are you going to do with that 25 watt LED?

<a href=showthread.php?s=&postid=8088055#post8088055 target=_blank>Originally posted</a> by pulse13
Speedstar, does the $250 you paid for the LEDs and power supply include optics?

Qwiv, I'm interested in using a PIC contoller for an LED project I'm working on. I'll probably be using 2 Lamina Ceramics Titan RGBLEDs driven by a Luxdrive buck puck for each color channel. Can your PIC controller give input into the buck puck for dimming? The RGB Titans are 25w LEDs with multiple red, green and blue chips in each one and seven 60 degree reflectors built in.

They only have a press release on their site but they are taking orders and I've already received a RGB Atlas LED to play around with.

www.laminaceramics.com

 

[pulse13]

I'm planning an LED hood for a 10g tank. I was planning on using Luxeons untill I discovered the Titan LEDs. I haven't decided for sure which LEDs to go with but I'm leaning towards the Lamina LEDs. I'm not sure if I like the color rendering of the RGBLEDs enough for tank use.

 

[WarrenG]

The light we see, whether it's from a MH, LED, or fluorescent is a combination/mixing of various peaks in the spectrum. Some amounts of 450-480nm, some amount of 550-600nm, 600-720nm, etc., all added together to produce the color temperature (Kelvin) that we see.

http://www.advancedaquarist.com/2002/2/aafeature

A MH light has various peaks in its spectral graph and corals respond to certain peaks and relative amounts of those peaks. How do the peaks and amounts of those peaks of the LED's compare to that of the MH, or various VHO/s and T5's?

 

[pulse13]

Click on the product number on this page to see spec sheets for each of the LED types:

http://www.laminaceramics.com/products/bl3000.aspx

White phosphor-based white LEDs like Luxeons and most others have a narrow blue peak centered at ~470nm and a broad peak from the phosphor centered in the yellow region of the spectrum but covering green through red. RGB LEDs have narrow red, green and blue peaks, I'm not sure if they have more energy in the red region which corresponds to chlorophyll absorption.

MH and fluorescents I believe are refered to as narrow bandwidth sources of light. The peaks in MH vary but are typically much sharper than fluorescents. There is also a lot of variation in fluorescent tubes depending on the particular phosphor types used.

White / blue LEDs lack output below 450nm whereas most aquarium MH bulbs and actinic fluorescents don't. I'd like to see some plots for aquarium fluorescents.

 

[WarrenG]

Looking at the link I did not find any plots of the spectrum graphs for their LED's.

Also, if the spectrum is very peaked, and is missing some colors, like yellow, or red, or purple, or orange, then that LED will not render that color very well.

Interesting info here...

http://www.wetwebmedia.com/ca/cav1i4/fluorescent_lamps/fluorescent_lamps.htm

 

[pulse13]

Oops I thought their datasheets had the spectrums, this link for the Luxeon K2 datasheet on p. 12 has the spectral plots:

http://www.lumileds.com/pdfs/DS51.pdf

Thats why I'm still debating using RGBLEDs, they have a red peak at 625nm which renders deep red well but not yellow-orange,orange or red-orange. Adding in red-orange LEDs ~595nm would help.

 

[WarrenG]

If a coral lives best at 30 feet deep in the ocean that coral is receiving almost no light from the red part of the spectrum. A bit of orange and yellow, yes, but not red. If it lives best at 15 feet deep the light the coral receives will have more red relative to blue and other colors than at deeper levels.