LED lighting

bdejong

Reefing newb
Hello all,

I am new to this board. I took a long hiatus from saltwater tanks (8 years). All I can say is WOW have things changed. I am setting up a small 30 gallon reef and right now it only has the water, substrate and about 30 lbs of live rock. I plan on adding and addition 30-40 lbs more this weekend.

Anyway, right now all I have is the standard 17w plain old top light that came with the tank. While it does light up the tank I know I need something else before I go farther.

It just so happens that I am an engineer for one of the largest led manufacturers in the US, We actually make the led chips. I have been tinkering with using LED lighting on my tank.

Right now I have a light fixture that uses only 20 of our super bright XLamp leds. These are extremely high power white LED's, in fact we recently completely replaced all the metal halide lamps in one of the city s parking garages. Not only does it use 40% less energy it is a lot brighter.

What I am wondering about is what kind of light is required to keep a good reef tank with all the nice inverts and coral? Is there a certain spectrum, color temperature, brightness minimum that I should be looking for? I can imagine that it is not enough to just have a lot of white light. So any information would be help full, I am trying to avoid using hot halide lamps and its a good DIY project.

Here is a link to the datasheet if anyone is interested,

www.etgtech.com/pdf/Xlamp/XLamp3_7090.pdf

Here is a picture of the new tank (6 days old now) with the single light using only 20 of the LEDs at minimum rated current. I am going to make another one and see how it looks. I also plan on making one with a few of the royal blues for a moonlight lamp and put them both on an alternating relay setup so that when one turns off the other turns on.

LEDTankLight1.jpg
 
Let remember that PAR is not a measurement of light, PAR as I know it is an acronym for Parabolic Aluminum Reflector. They are used for metal halides since they are a round bulb and if you cross section it only 180 degrees of the bulb is directed to the tank, the reflector is used to bounce the other in the direction of the tank. This is where the LED should shine (pardon the pun). The LED is a forward directed device, while most T5 (the little round ones most people are familiar with) have a very narrow beam pattern and as such you would need a ton of them to provide a full spread to the tank. The XLamp device on the other hand has a 100 degree spread and is very efficient.

There is also the benefit of adding either blue or red or orange leds with the white ones to skew the resulting spectrum to mimic the type of light you desire.

Lets think about this as well. Since fish can not close their eyes their sleep pattern is dictated by us. Lets push this one step further, how would we feel if our sun operated on a on or off schedule. Picture being on the 18t tee, its getting toward the end of the day, you place the ball on the tee, get ready and half way through your swing the lights go off. I know my stress level would be through the roof. The sun doesn't work that way, it gradually rises and falls below the horizon. the tank lights should do that as well. Moon lights are not the same, it is still going from 12,000 watts of light to a few watts in a blink of an eye.

What I propose to do is program the lights to slowly transition from full to moon over a period of hours like the real light does. This is very possible using LED's, they can go from full power to minimal with no fall off in color.

Another thought, what would be cooler than a light that would allow you to dial in extra Blue or red, or yellow, or orange. Its your choice depending on the tanks needs. Want the blue hue of an Actinics then turn up the blue, ant the color to stimulate the plant growth then turn up the reds and oranges, throw in a little yellow and you can be all.

This is very possible using a bank of Whites with additional colors with individual control.

What do you think????
 
They do sell LED's for aquarium.Solaris LED Hood PFO lighting Inc > Home ( DNN 4.0.3 )

I read up on this type of lighting all over the web and aquarium magazines.It would seem that the jury is still out on them.

12,000 watts to a few watts?You was kidding right.I never heard of anybody running 12,000 watts.Actually dust to dawn is the transition from lunar to actinic(sunrise) to daylight then back to actinic then lunar lights.Its not as drastic as you make it out to be.

I am curious about how this turns out for you though.
Keep us posted.

Personally to keep any coral or clams Mh's are the way to go.
Some say if your tank is shallow enough,oh 20inches or less than VHO's or T-5's with individual reflectors should work.
 
Last edited:
I have thought about trying to set my lights up to gradually brighten and dim to simulate dawn to dusk.Even thought about trying to get one of those solaris fixtures.But at 2300 bucks that would have been a sure divorce.
I do how think that even using actinics to simulate dusk and dawn is not natural,I dont think it causes and stress to our fish or corals,but its not natural.Spend a morning or evening and watch how the sun rises or sets and youll notice that if very gradual.
Just my 2 cents worth
 
They beat me to it. I had not heard about the solaris, but yes that was the idea. $2300 would indeed be a divorce purchase, even the 800 small unit would cause several nights on the couch.

12,000 watts, I read someones post that quotes 12,000 watts. Could be overstating, could be real. :shock: I was only trying to get the point across that he sudden transition could be a point of stress, maybe not enough to cause harm but hey anything we can to to make our pets happier right :bounce:

What is it about MH's that make them the best? I have been trying to find that information but I can seem to. I think they are popular for our hobby for the same reason they are popular for street lights and large area lighting, they put out a lot of light for their small compact size, and once started they are efficient. They are available in several color temperatures, for those that don't know that is the 8000k to 12,000k number.

If the same amount of light (brightness) can be achieved at the same color temperature then I don't know why LED's couldn't be a direct replacement. I can tell you that it is absolutely possible with LED's to get the same color temperature and brightness of a MH bulb with high power LED's. I can also tell you that it is possible to do this using less power.

Here are some other interesting article that you may wish to read.

http://www.redherring.com/Article.aspx?a=21087&hed=Could+California+Ban+the+Bulb%3f

Photos: LEDs light the way in Raleigh | CNET News.com


The idea of variable color temperature is as simple as adding colored LED's to the mix of whites. Even the whites can come in several different color temperatures.

Add the cost saving and reduction in heat generated it is only a matter of time before these are mainstream.

I will keep you all posted on how it is working. This is a definite project for me and I already plan on measuring several MH bulbs in the lab next week to get a good baseline for their replacements. Since we have already done a lot of this work before for other projects I have a good idea as to what it will take to get the job done right.

Thanks
Brian
 
I think that that th MHs put out more lumens or something like that.

Yup, they put out a ton more light for their size compared to standard compact fluorescent tubes. From what I can gather that is the main reason they have become popular. Compact size, lots of light, and basic housing. To me the only downsize is cost and heat. Please do not take that comment as a digg against the MH lamps, it is not. In everything we do there are pros and cons. The above are some of the Pros and a con, the LED solution will have its pros and cons as well. The biggest con will be the initial cost,but that will hopefully be earned back by the pros of less heat and less energy usage. As noted in the Parking garage installation they have found a 35-40% decrease in energy usage and they predict it taking only a few years to recover the cost of the install, after that its all savings. The lifetime rating of the LED is 50,000 hours, thats 13.5 years running at 10 hours a day. A definite Pro.


Here is what the first fixture looks like, very simple construction.

_MG_7890.jpg
 
MHs are considered the best I think is its ability to penetrate the deepest parts of the reeftank.I definitely not an expect on MHs so hopefully a moderator here may jump in with their experiences.

What I read on another forums,some moderators test these LEDs and their coral didn't open well under them.I know some coral may open alot when they are not given enough light,so I'm not certain that this was the case.Their final statements were that they felt that LEDs didn't penetrate the lowest part of their reeftanks like medal hallides and that it would take more LED bulbs then a standard size fixture to accomplish comparable results to medal halides.

I curious to know how you incorporate the slow transition from dusk to dawn.In my mine,that would seem like a gazillion timers.Is their an alternative to the bulky outlet kind and the digital ones that can transition the amount of LEDs needed for a more realistic day cycle?
 
What I read on another forums,some moderators test these LEDs and their coral didn't open well under them.I know some coral may open alot when they are not given enough light,so I'm not certain that this was the case.Their final statements were that they felt that LEDs didn't penetrate the lowest part of their reeftanks like medal hallides and that it would take more LED bulbs then a standard size fixture to accomplish comparable results to medal halides.

Interesting, this is the kind of information that I am looking for. Here is what I think is going on. Most, if not all commercial white LED's target the standard household color temperatures. This is usually from 4500K to 6500K. this is what "people" want to see so that is where they are targeted. What is missing form all of the LED lamps I have looked at so far is additional BLUE (450nm) LEDs.

Advancement in the Lumens per watt output of a power LED is happening extremely fast right now. Consider the normal cycle for getting a product into production, the devices that they started with 6 to 9 months ago are now considered low output.

I would think that the main reason some of the early test lamps didnt do to well was from a combination of not enough light output and not enough light in the right wavelengths. What I am finding out is that the MH lamps that are popular are very heavy in the blue spectrum. The appearance of these lamps is noticeably blue. In some cases very blue. It would be unacceptable for most all LED applications to produce a "white" LED with such a high color temperature.

The next step for me is to do some comparisons in the lab of the typical output of a MH lamp and then test how many LED's it takes to produce the same amount of light. Take a look at the following link, it is a picture of the new light fixture used in the parking garage. this one fixture replaced a standard sodium lamp and puts out more light with less power.

Raleigh, N.C. to replace street lights with LEDs | Planetizen Radar

I curious to know how you incorporate the slow transition from dusk to dawn.In my mine,that would seem like a gazillion timers.Is their an alternative to the bulky outlet kind and the digital ones that can transition the amount of LEDs needed for a more realistic day cycle?

This is another area where using LED's is easy, its as simple as varying the duty cycle to the LED's. Powering the LED using a simple PWM controller is not that hard, and will enable the dimming of the output. What I dont know is at what point the LEDS color temperature starts to fluctuate. But I will find out.


EXCELLENT discussion, keep them coming. I have tried to post a picture twice now but the post is not showing up. I will try again here.

_MG_7890.jpg


Brian
 
reeffreak said:
I curious to know how you incorporate the slow transition from dusk to dawn.In my mine,that would seem like a gazillion timers.Is their an alternative to the bulky outlet kind and the digital ones that can transition the amount of LEDs needed for a more realistic day cycle?

Im not sure how bdejong would implement it but i would try to desing a timer with a pulse width modulator that would change the persent of time the LED is on versus off. Lets say during the first hours, it would tell the LED to be on for 10% of the time. This would be like 1 millisecond on 9 milliseconds off. then the next hour bump it up to like 30% and so on untill it is on 100% for the peak mid day hours then gradually go back down until only the moon lights are on.

Brian
 
(Read a few posts back, I talk about using a PWM controller)

On the transitioning of the light it would be just like described, using a PWM controller (basic in even the smallest micro controllers) I could duty cycle the leds from the main ones to the moon lights in a nice smooth transition.

Here is a screen shot of the spectral response as the light sits right now. It is reading at about 6700K. I am working on the supplementary BLUE LED bank right now and hope to have it installed in a few days, This should bring up the color temperature as well as the contribution to the blue portion of the spectrum. I am also looking into getting a set of higher color temperature White LEDS to start out with.

This is the baseline so that we can compare the light as I make changes. It should be interesting.

SpectraResp2.JPG
 
Great link, thanks. I am using the same equipment, although a better spectrometer (HR2000-High resolution instead of the standard SR2000). I am hoping to spend some time in the lab this weekend installing the complimentary blue LED's. :lamp5:

I do have to take exception to their statement that using the Philips LED was a good choice. The XLamp would have been a better one hands down. (Read a little company pride into that statement) :mrgreen:

Think about it though, a light with only 25 devices was fairly equal to the 250 watt MH bulb. Only using 75 watts of power its a fairly large savings not only in heat generation (which is wasted energy) but also in power consumption. While power consumption is not a huge factor on a small tank like mine, I can image that it could be when you start looking at a 150-200 gallon tank where you are using several of the big daddy 400 watt MH lamps.

I will post more later, for now I have to get to work.
 
I'm fascinated how this works out for you.Always like the shimmer of MHs but not the heat or electricity consumption.,since my limited expertise and income I can't be an experimentalist.

keep us updated on your progress and good luck.
 
Here is your update. I added 10 Royal Blue LED's to the mix to bring up the color temperature. The measured color temp is approx 17,500K. This is an average since at certain measurement points the spectrometer could not calculate the CCT.

As you can see the blue is peaked now. The color in the tank is great, not over blue but definitely more white.

The extra blues had to be mounted on the sides, so I am expecting that when I make the new housing I will be able to remove a few of them to bring the power consumption down.

All of the LEDs are running at only 350ma each. The light housing is mildly warm to the touch and is mounted about 4 inches above the water. There is no heat transfered to the water. Since reading the review article I am now monitoring the temperature differential from light on to light off. I have yet to notice more than a degree difference.

SpectraRespExtraBlue.JPG
 
Back
Top