DIY Led 10 gallon build thread/alternative diy co2 thread

[krap101]

Thanks! It think it might have been in your thread, but what I believe was discussed was leds just in parallel (like 3.7v ps with high amperage, and running 1 led in each string.) I'm not familiar with the meanwell, but I have taken E & M in college, but applying it to real life like this is just funky as it should be.

I'm curious to know what would stop you from running a parallel series on a meanwell. It might be something inherent in the design, but from my E&M knowledge, splitting the string should double the amount of current and halve the amount of voltage. Like I said, I'm not sure if there's something inherent in the meanwell that will prevent me from doing this, but I thought what I learned could be applied to any power supply.

Also, I probably won't be running more than 6 for this project, but I'd like to be able to use the same design, and maybe the same meanwell when I upgrade to 29 and 40B, which is why I'm trying to future-proof the design now.

Since I'm using a driver, wouldn't that prevent most fluctuations within each string, since current within a string is the same for all "resistors"? I'm trying to find out what instability is there when using a parallel (or parallel series) circuit


Thanks so much guys, sorry for the essay I wrote . I've found that with the delay of forums, it's not like iming, where it's best to get as much information out as possible

Thanks again

 

[mrg02d]

The problem with parallel strings is that you cannot get the rated current through both strings. With two parallel strings, you will get half. This is because the meanwell will keep the current drawn at 1A (for a 1A model).

So if we have a 1A current source, and two parallel strings then we will have 0.5A go through each string. The current will divide at the node connecting the two strings to the positive terminal of the meanwell.

There is also the method by which the meanwell senses the current draw and adjusts itself. Many common constant current sources have a way to detect, in real time, the current drawn. They then adjust to maintain the predetermined current. This usually is a resistor that they measure the voltage across and via ohms law (v=IR) they know the current being drawn. I built drivers that would output 75v and control 60A. These used and op-amp and current sense resistor for each channel. The channels could be put in parallel for more current, but still had a way to detect the current through each channel to keep it constant. Id check with meanwell to see what they say.

0.5A + 0.5A = 1.0A

If you add a third string, that will then be 1/3 the rated current.

This might not be what you want, but its what you want the meanwell to do. (Keep the current constant regardless of what you or the line power does).

This may benefit you if your lights are already too intense. (A way of dimming them).

I suggested 16 simply to avoid spot lighting. Optics would increase spot lighting as well. My light has the leds spaced 2in apart and the strings spaced 3in apart. It covers my 12in^2 area nicely, but just so.

I tried it over my 20g reef and was very dissapointed. Its much larger, but the spread was terrible. My 20g is 13in by 24in. It looked like it would need three of my lights AND optics.8-O

Deffinately will leave the t5 over the 20g.

We dont add leds for the watts/gallon thing, but for even spread. We dim the leds for the watts/gallon. (usually dim them because they are too intense). This is why the led lights are costly. They need a lot of leds to cover the tank well.

I wouldnt concentrate so much on the idea of total power for two reasons:
1-LEDs are inefficient. They draw 48w but they sure dont put out 48w of light. Same with t5 and MH. Youll have to look up your led efficiency.
2-Nothing in the tank will see the total combined light output. The sun is a great example. I forget how much power 1m^2 of sunlight has, but 1in^2 within that 1m^2 certainly isnt equal.

Matt

 

[Jimbo7]

Real quick for a visualization. 3x4 pattern, lights 8" off the tank. You had said 14" but that seemed really super high to me. Any reason in particular you are going with 14"?

White is 90 degrees viewing angle, Blue is my estimated range of max power, 60 degrees.

 

[Jimbo7]

I wouldnt concentrate so much on the idea of total power for two reasons:
1-LEDs are inefficient. They draw 48w but they sure dont put out 48w of light. Same with t5 and MH. Youll have to look up your led efficiency.

Matt

What do you mean by this?

 

[mrg02d]

Thanks! It think it might have been in your thread, but what I believe was discussed was leds just in parallel (like 3.7v ps with high amperage, and running 1 led in each string.) I'm not familiar with the meanwell, but I have taken E & M in college, but applying it to real life like this is just funky as it should be.

I'm curious to know what would stop you from running a parallel series on a meanwell. It might be something inherent in the design, but from my E&M knowledge, splitting the string should double the amount of current and halve the amount of voltage. Like I said, I'm not sure if there's something inherent in the meanwell that will prevent me from doing this, but I thought what I learned could be applied to any power supply.

Also, I probably won't be running more than 6 for this project, but I'd like to be able to use the same design, and maybe the same meanwell when I upgrade to 29 and 40B, which is why I'm trying to future-proof the design now.

Since I'm using a driver, wouldn't that prevent most fluctuations within each string, since current within a string is the same for all "resistors"? I'm trying to find out what instability is there when using a parallel (or parallel series) circuit


Thanks so much guys, sorry for the essay I wrote . I've found that with the delay of forums, it's not like iming, where it's best to get as much information out as possible

Thanks again

Not sure what you mean by splitting the string, but...

Loads in parallel will have the same voltage across them, and will attempt to draw more current. You can reason it in that more parallel paths means more routes for current to go. The more parallel strings you add, the more places for current to go. The voltage stays the same across each parallel string.

ANother way to view things is peer pressure. The voltage is the peer pressure, and the parallel paths are more things the peer pressure wants you to try. (Smoke this dope, kill this bottle of vodka, and then snort this stuff). Three no-nos and peer pressure mean three chances to get in trouble. The more bad things to try, the more chance to get into trouble. You may resist well under the current peer pressure, but once it goes up (voltage rise) you may crack and give in. Unless you have a constant current controller.

Not sure what E and M is, but this comes from Kirchoffs current and voltage laws via nodal analysis and loop analysis.
That would be from my EE classes.

Matt

 

[mrg02d]

What do you mean by this?

The led will draw 3.7v at 1A (or so) and that is 3.7W. It will NOT put out 3.7W as that violates the laws of physics. It will put out: power in times efficiency. So if we had a 50% conversion efficiency, 3.7W in would mean 3.7/2W out of light. So a light consuming 48W will only produce 24W of light. (Under ideal circumstances).

Well, off to my class...

Matt

 

[Jimbo7]

But how many watts of light it puts out isn't as important i thought. What matters is the lumens and PAR for our uses.

 

[Crepe]

Aside from the light issues I'm thinking about discussing your CO2 method. Unless you can control the temperature of the thermoelectric plate there's no controlling your co2 output throughout the day. The Sodium Bicarbonate will decay at a rate non-linear to temperature due to the fact that rate of reaction will be dependent on the mass not only temperature of the substance. Your rate of reaction constant K in NaHCO3---> NaO2 + CO2 would be constantly changing. Furthermore your CO2 output is minuscule in comparison the amount of NaHCO3 you need.

Let's do this:
1mole NaHCO3 decommposes to 1 mole Co2
While accounting for experimental loss it's close enough to 84 grams of NaHCO3 to 22.4 litres of gas at STP. At 50C [temperature of decomposition of sodium bicarb] or 323.15K.
Using the charle's laws: V/t=v/t. 22.4 litres will become a whopping!:...26 litres of gas.
Now 26 litres of gas may seem like a lot, especially out of only 84 grams (about palmful) of NaHCO3 but imagine the difficulty of controlling the gas. Once it hits the temperature of reaction, it's not going to slow down, instead speed up then slow down when only a minuscule amount of bicarb remains. I can't find the rate constant to show you the exact rate of reaction but considering it only took a few extra seconds for me to turn a perfectly good anhydrous sample of CuSo4 to copper and sulfur, it's not very long.

Stick with a traditional diy co2 and also a traditional aluminum heat sink. If you really want the ultimate in cooling pick up some Arctic Cool Silver Thermal CPU compound and buy a Zalman all copper heat sink.


All in all, bad idea.

 

[krap101]

Well, baking powder (sodium bicarbonate with a acid in solid form) releases its co2 when it touches water (water dissolves whatever is keeping the acid from reacting, so you get sodium bicarbonate + acid=co2 + something else. I was thinking you could use something akin to an iv bag and drip water in at a consistent rate. I'm worried about the used up baking powder compacting, and eventually stopping the water from reaching the rest of the unused baking powder. I'm not that familiar with baking powder, so it may not act this way.

I know that watts are a bad form of measurement, but it is hard to measure par (expensive or rare maybe)

I mean splitting the string as in rather than putting all the leds on one string, add a second string. You'll need half the voltage and twice the current in this case.

An led either uses more than the rated wattage to put out the rated amount of light, or it uses that much energy and puts out less light. Yes they're not 100% efficient (they do put off a decent amount of heat when at these specs) but compared to fluorescent metal halide incandescents, it is a drastic increase. I think the last time I checked were at around 70-75% and were getting closer to 80%?

For the parallel strings, you're saying that I can do it, just it has a greater chance of risk? If I get the 24V, I'll have around 2.5A of current I can play with, which is around 3 strings running the leds at 800ma. I'm unsure of what'll happen to the extra .1A. will it just not get drawn through, or will I have to dissipate it somehow?

My 10 gallon is 12" high, so with no gravel or anything, the bottom of my plants will be around 14" below my lights (2 inches above frame/water)

 

[Crepe]

You need a tiny amount of baking soda solution to reach 30ppm and even then adding excess Na+ ions from bicarb is bad for both plants and fish. 30ppm co2 makes 30ppm Na. You should never have that high of sodium in a fw tank. And this doesn't account for the fact you have to add a continuous supply of bicarb to maintain 30ppm. That 30ppm sodium is just the initial dose. After a day you may have a few hundred ppm na.

 

[krap101]

I'm not adding the bicarb to the tank. It's actually similar to a yeast generator, but instead of using yeast, you use baking powder. In order to control rate, you control how much water is added to the reactor. You'll only need to add more bicarb (change the "reactants" out) when the baking powder is used up. It's similar to the yeast reactor this way too, except you don't have to worry about keeping the yeast alive (or anything for that matter), I'm guessing it will last longer, and you will have more control over how much co2 is released (how much water is dripped into the powder.)

 

[Crepe]

Decomp of H2CO3 is very miniscule. You aren't going to get sufficient co2. The amount that comes out of the solution won't be enough. You are not only fighting the bonds of the H and Co3 but also the partial pressure of the actual solution(which is lowered) and the atmospheric pressure. In short, it probably won't work sufficiently well. EI: the combined pressure of built up gas against the water and the high humidity and high concentrations will all combine to equillibriate at an undesirable point.

 

[krap101]

Baking powder is baking soda + dry acid. When you add water, you release the acid, so you will get 1 mole co2 per mole of baking soda. This reaction will be done at room temperature, or maybe higher if I want to heat it with my light.

If I want to be able to control the reaction even more, I could start using vinegar or some other acid instead of baking powder, and dilute it until I get the right amount of co2?

I'm guessing i'll need something to absorb the extra water, so maybe use some bentonite clay (well digger's mud) or a sponge at the top or bottom of the reactor.

 

[Crepe]

Acid is probably your best best. A weak dilution of vinegar is going to generate the most co2 consistantly. Only problem now is convenience. How often are you going to refuel it? a 20# tank of co2 lasts 6months-year on a 10gal. Yeast lasts about 1 month- 2 months.

 

[krap101]

It depends on the byproducts of the reaction. I think you'll be able to create a very long lasting reactor with this method, just you need to find a way to control the byproducts. I believe that you just get a salt, and if it's water soluble, I may be able to get rid of it using the sponge/bentonite, The bentonite is used for oil spills in shops and when kids throw up, and I'm hoping it will absorb all the excess liquid after the reaction takes place, and hopefully absorb the salt (sodium acetate). If I can keep the contact point between the new vinegar and the baking soda free of salt and water, I think it will increase my chances of success by alot. And I'll be able to continue to add vinegar until the baking soda is all used up, or the contact point is covered by sodium acetate.

I wasn't aware yeast lasted that long. 2-4 weeks are the numbers I'm used to.

 

[krap101]

Looking at some leds now,

From this site
http://www.cree.com/products/pdf/xlamp7090xr-e_b&l.pdf

and the graph, that means I want a WC, WF, or WN right? 6700K is the optimum temperature?

 

[Jimbo7]

You probably want to go with XR-E's
http://ledsupply.com/docs/cree-xre.pdf
I think a mix of cool white and neutral white would work.

 

[krap101]

Since I can't edit my posts in this forum, here are more links

Designing and Building a LED Fixture
Ultimate LED guide - Nano-Reef.com Forums
Comprehensive DIY LED Project List - Nano-Reef.com Forums
Cheap LED Light?
http://www.aquaticplantcentral.com/...led-luminaire-mkii-may-2010-a.html#post547327

As of right now, I'm probably going with the cree xre p4/q5's. I'm thinking neutral white is a little bit too yellow? Or maybe not..

 

[krap101]

LED Experimental Tank (to be started later) - Page 7 - DIY Aquarium Projects - Aquatic Plant Central

I think I found my guideline. It isn't a wpg guideline, but rather a sipled (square inches per led) guideline.

Sorry I've been gone, But regarding how many LEDs you need for a tank, a good rule of thumb I've devised is for low-medium light, 1 LED per 24 square inches of surface area of the tank. I.E. my 29g has 15 LEDs over with the top having a surface area of 360 square inches. Higher light would be in the 1:18^2" or for insane light 1:12^2" ratio.

Do not attempt to use Lumens as a rule of thumb, as LEDs get better at emitting light, keep in mind PAR levels may or may not jump.

Evil, I do run just 6500K, I'm not sure I want to run anything else, nor am I sure that anything else is required. I think the red is hyped a little bit much, most people run 6500K PCs without much issue, and I think that a case could be made that LEDs put out more reds and blues than their fluorescent counterparts.

That said, I am beginning to think that the planted aquarium could stand a little bit more red to help in the root system (most reds are used to help drive the root system from what I've red). I may or may not add a few reds to try and prove that theory.

Andy..... Its really hard to say, the 1:24^2 inches ratio is deff not medium, not when I'm dosing ferts every other day or so, with DIY CO2, in a 29g tank, where 2 bottles isn't giving me 30PPM. I suspect its borderline high light, medium would probably be 1:30^2"-1:36^2", its really hard to say without a PAR meter. 60 is deff in the high range though, going higher without CO2 is not adviseable.

Mileage may vary by tank depth of course, my tank is 17-18 deep.

With all that said and done, anyone the Greenville/Spartanburg area with a PAR meter be willing to lend it to me for a day or two get some numbers on my rig?

Currently, I'm planning on 8-10 leds. I can do 2 strings of 4/5 or 3 strings of 3. As the tank is 20x10, having a 2x4 string setup sounds like it will have the best "uniformity". So, using p4's, that will probably cost around $28-$35. I haven't decided on how I'm going to heatsink it yet, either one big one or a few small ones.

Not 100% on power supply either...yet

I'm also wondering, if the cool white and neutral white run the same voltage, can I put them on the same string as the cool whites? I'm thinking maybe 20-30% neutrals now that I've seen some tanks with that on them.

 

[taylorodw]

What are the advantages of LEDs in a FW system? How much does this look to end up costing?

Do you see, at some point in the future, an "idiot's guide" type thing? IE, how anyone can learn the cycle even if they don't understand the science behind it... will LED aquarium lighting get to the "do x, y, z, and you're good to go" level any time in the near future?