LED selection for FRESHwater aquarium lighting

[artemm]

Hello all!

I am new to led aquarium lighting but i am pretty experienced electronics hobbyist. So, for me is not problem to build a smart and controlled DUY lighting led system.

However, there are question on what leds to choose and how to light them up.

Let's start from the simple one

1) Methods of lighting

There basically two way to light up a led: PWM and constant current. With PWM the leds are brightly pulsed with some frequency. So, if F is 100hz and over it looks like constant light to out eye. If we make pulses shorted (for example, duty cycle of 10% on and 90% off, frequency is the same) the leds look dimmer. It is very easy to control the brightness of the leds this way and easy to do rising and setting sun effects as well as other effects. The constant current is a more sophisticated way. It actually generates constant flow of current through the leds and control the amount of this current. No pulses here.
For PWM i can do about 3-10KHZ frequency without any problems.
Now, the question is what will plants do about it? Is PWM bad for them? Will they react normally for PWM lighting methods or they need constant current way only? Or PWM is okay starting from some frequency and up?
I have not found any answer to these question anywhere. Any scientifically backed up idea from anyone?
Actually, the PWM pulses can delivery a lot higher momentary energy density than the constant current method. So, i don't the the details of the chemistry, but PWM might work worse or better that constant current method.
Anyhow, if no answer here i will just built a constant current driver or buy a chip for it.

2) LED light Spectrum

This one a lot more important. I am building this for freshwater aquarium to light freashwater plants (and if it work i will try it for usual pot plants too). So, according to this:
http://www.americanaquariumproducts...._Lighting.html
and wikipedia also (see PAR)
i mainly need two spikes: around 470 nm and around 670nm, avoid most yellow and green and avoid UVA,UVB. Better yet have a good plato covering
all from 620 to 700nm.

There we also see that they suggest white leds with hight kelvin color temperature.

But let's look at the pdf for XP-E leds from Cree.
http://www.cree.com/products/pdf/XLampXP-E.pdf
Page 6
we see the power distribution for the wavelengths.
As we see there WARM WHITE (lower color temperature) has actually a lot more energy in 620-700 nm range than cool white and its power distribution falls a lot better into 620-700 range too. The blues are lower, but we don't need much of them too. The middle section is pretty much the same for all whites.

Then why high color temp white leds are suggested if according to the graph
warm white fits better?

Also, isn't it logical to add a bit more red leds with middle spike around 670nm to add more energy in 620-700 nm range, as well as add some with 480nm spike to increase photosynthetic response? Let's say, 1 extra red led for 4 white and 1 extra blue for each 6 whites.


And one extra question: i can really put an automated scheduled UV-C range sterilizer. Is it true that it will kill the algae?

 

[BigJim]

1. I've seen reef tanks and planted tanks running well using LEDs operated on PWM control. Plants are less picky about light than coral. LEDs put out good PAR and that's all the plants really need. I believe I have some Cree XR-E LEDs coming for Christmas that I'm planning on using with PWM on my tanks.

2. I read through that same link. It's pretty informative. High-power LEDs in the 6700K spectrum seem to be the way to go. They look good while providing the light spectrum your plants need. Warm white colors can give your tank a nasty yellow cast. One of the beauties of LEDs are that you can mix and match until you've got a spectrum that you like. I asked for the cool white XR-Es, but I may add some warm white or maybe even red LEDs to round out the spectrum.

An effective UV sterilizer will kill all water-borne algae. It won't really affect the stuff on the sides of the tank.

 

[artemm]

1. I remembered, that Fluorescent lamps were used forplants lighting for decades. And even 10 years ago majority were lit with old ballasts which flicker at 100hz. And they even grew vegetables with such light. So, this is another confirmation that PWM is fine for plants. However, i don't know how to measure PAR output of PWM lighting because PAR meter can lie badly with PWM. It depends how it accumulates the reading and how fast is the sensor.

2. 6700K leds by themslevs do not provide enough power in red part of the spectrum which is a lot more important for plants than blue part. They simple put most of its power to blue part. If you look at the pdf i linked page 6 you will see tha the middle part seems to be fine and a bit yellow is added in warm white because the orange-red range is stronger, but the spectrum for warm white a lot better for plants. Do you angree that warm white leds provide a better spectrum for plants? Let's not mention algae and other things for this moment, consider only the plants.

3. Really? I never had any noticable quantity of water born algae, but i though it spread though water, so, if it gets kill in the water then nothing will be on the glass, plants and stones. I am wrong?

 

[BigJim]

I have no experience measuring PAR, but it's always important to have the right measuring device for the job.

I have seen PAR readings on an LED reef tank build for both the original lighting (T5HO, VHO, etc.) and for PWM controlled LEDs and the LEDs were better at all depths. I don't have the link handy, but it had pictures of the tank with the PAR readings at the location they were taken. It was a pretty comprehensive study.

From what I've read, the blue light is important for growth, the red light is important for flowering and propagation. I picked the 6700K LEDs because I like the look on the tank and they have an acceptable spectrum. I've found that different 6700K lights can have hugely different spectrum analyses. Good plant lights put out light in the red and blue spectrum that averages out to 6700K. Cheap lights put out light in more of the yellow-green spectrum that averages out to 6700K. At the moment, I have cheap T12s, spiral CFLs, and good T5 lights on different tanks. I can grow plants more easily with less algae under the T5 lights with less wpg than the T12s. The CFLs fall in the middle with better results on shallower tanks.

The sterilizer would take care of any algae spores (or whatever they use) that pass through it, but I don't think it would magically make the algae in the tank disappear.

 

[artemm]

It would be just great if you could find the link that study.

For the algae issue. The question actually appeared because with better light the possibility of algae rises (even with the correct spectrum), so, i must either add CO2 to boost usage of the nutrients by plants or control algae some other way. So, it is not about making algae disappear, it is about preventing from appear in the first place.

The question is how to calculate how many leds i need for a pacticular setup? So, i would at least know how many order.

Has anyone seen a table with number like PAR, water depth, halfangle of the led, led model + distribution of PAR vs distance from the center of a light spot.

 

[artemm]

For sterilizer i meant to to get rid of algae but would it stop from appearing. I have none now, but it might start growing when leds are in place.

How to calculate how many leds i need for a pacticular setup? So, i would at least know how many order

Has anyone seen a table with number like PAR, water depth, halfangle of the led, led model + distribution of PAR vs distance from the center of a light spot.

 

[BigJim]

PC PAR - 260watts (same picture used for LED PAR)

LED PAR

DIY LEDs - The write-up - Reef Central Online Community

Here's one of the articles I found. The author compares PC lighting to his DIY LEDs.

I haven't found any tables or rules of thumb for LED aquarium lighting. It's really difficult to make any direct correlations between LED and other lights. LEDs are very directional and don't have restrike problems. The thread I linked to uses 24 white and 24 blue Cree LEDs over a 75g reef tank and generated good PAR readings. That thread says the LEDs without the aid of optics are just short of a 400W MH light and they put out light with better spectral peaks according to the link you posted before.

This might be the best empirical data you're going to find. I'd start with this and modify accordingly. If I get my LEDs for Christmas, I'm probably going to start with 5-6 over a 20H to see what they can do.

 

[Ziggy953]

Art, Have a look at my thread. I'm using the same design as mentioned above to light my tank. I have a 54g corner tank that is heavily planted and is lit with 24 LEDs. link is in my sig.