Been a while...but my experiments continue!! 4'Degree tank!

[dannypd]

Alright, tis been a while since I made a poste, however, I am back...for the day...with my latest in "inventions:

-100% L.E.D. tank (no plants will grow, but it's very bright)
-4'Celcius goldfish tank (indoor wintering tank)

----

The first one I did (LEDs) was because of the lighting in my tank was not the brightest out there, but it did get fairly warm during the day, causing a stir in my tempurature and wasting more electricity.  I choose the closest to marine lights as possible.  The tanks lighting is powered by a 12V car "charger" battery (it's only half an amp).  That battery is charged via a 12V solar cell at 4.8W (1 day charge, 4 days light).   In the picture (attached to the next messages, if possible), you see a lamp with a "twist" type florescent lamp, the tank at the bottom is another florescent lamp "tube" (daylight) and finally the main attraction, marine blue L.E.Ds.

----

Second experiment:  I wanted to keep 15 goldfish (3-6 inches) in a 20 gallon tank with 25-30 gallon sump...in other words, bad....but, I did anyway, I just made the tempurature 4-5 degrees celcius.   The fish stop doing ... well... everything.  They will sometimes swim in an odd way, then huddle up together and stop again....but they do not eat, or excreat anything! yeah!   In the pictures attached (if they work), you can see the (unfortunate) frost/condensation on the fish tank, and the huddled fish.  In the other picture you see my "cheapo" chiller, a hacked dehumidifier (irony is the tank became the dehumidifier).  Also, to be even cheaper, fresh snow is generally placed into the tank instead of leaving the machine on.

[dannypd]

Pictures...

[darkdep]

VERY VERY VERY interested in your LED experiment...how many physical LED's do you have running in there?  Any thoughts of using "white" LED's instead of blue?

I LOVE the idea of using solar power (plus battery) to power the lights...

[dannypd]

VERY VERY VERY interested in your LED experiment...how many physical LED's do you have running in there?  Any thoughts of using "white" LED's instead of blue?

I LOVE the idea of using solar power (plus battery) to power the lights...

I am pretty sure I used 24 or 48 blue (not officially blue) LEDs, using my exact design (so, 3 per board, in series, using a 800ohm resistor.  The led's are rated for 0.20ma each.

Here, this is my pcb layout I created...  I hope to fill my entire house like this, so if you know someone that can mass produce my pcb board (just the layout/chip is enough), it'd be awesome!!

[darkdep]

I cannot tell you how much this interests me.  How much draw do each of the LED's take?  Do you have pics of your finished product (the boards themselves, not the light effect)?

Any guess as to how much light (in lumens?) this setup produces?

[dannypd]

I cannot tell you how much this interests me.  How much draw do each of the LED's take?

Not much...lol, it's 12 volts, and .020amps.  I am using a 800ohm resister for 3 LEDs.  I am not too good at electronic calculations.

Do you have pics of your finished product (the boards themselves, not the light effect)?

It's very disgusting and I am worried about opening the setup just now.  It's "airtight" because of the salts from my fish tank, generally, my lights are covered in it, so I prevented that this time around.  The head created from the LEDs (VERY minimal, but since it is airtight...) is dissapated via a few copper foils petruding the side of the lamp fixture (not going into the water at all...no worries there).

Any guess as to how much light (in lumens?) this setup produces?

Oh god no...lol...if I had to guess, maybe 15 lumens per LED, but that's a field of no more then 15 degrees.

[jdx]

That's an interesting idea indeed, I think you may get better effficiency by putting LEDs in serial and reducing the resistor's resistance or simply removing the resistor if you have enough LEDs in serial. If the LEDs are silicon diode, the diode voltage will be 0.6-0.7V, so if you put ~20 of them in serial, and supply 12 volt directly, you will get close to 100% efficiency on the LEDs. From you current drawing, the resistors actually consume majority of the energy.

[dannypd]

That's an interesting idea indeed, I think you may get better effficiency by putting LEDs in serial and reducing the resistor's resistance or simply removing the resistor if you have enough LEDs in serial. If the LEDs are silicon diode, the diode voltage will be 0.6-0.7V, so if you put ~20 of them in serial, and supply 12 volt directly, you will get close to 100% efficiency on the LEDs. From you current drawing, the resistors actually consume majority of the energy.

:shock: yeah, I so do not know electronics enough yet. lol.

Actually, I only discovered how to properly use resistors recently, because of fluxes in the power input from the sun, can be high, or low (battery takes over when it drops below 12volts).  Also, what if one LED goes out??  do they all die then?

p.s. how much of an efficiency drop are we talking, .0025 watts? lol

[jdx]

From a quick calculation:

From the drawing, the current through the resistor is:
I = (12-0.7)/800 = 0.014A

The power that resistor is consuming is:
P = U*I = (12-0.7) * 0.014 = 0.16 watt.

The power that 3 LEDs are consuming is:
P = U*I = 0.7 * 0.014 = 0.01 watt.

So the efficiency is only 0.01/( 0.01+0.16) =  6%, there is a chance to improve efficiency by more than 10 times if LEDs are put into serial.

Yes when one diode dies, it normals opens instead of shorts, the whole chain won't light but it won't burn the rest of the LEDs, even if it shorts, it's equivalent to increase of 0.7 volt to all the remaining LEDs, the LEDs can tolerate a variation in voltage so it's not a big problem.

[jdx]

The above calculation is base on assumption of a forward voltage of 0.7 Volt, for some LEDs, the forward voltage may differ, I think some are as high as 3-4 Volt.

[darkdep]

Ok, so before I brush the cobwebs off my electronics knowledge, can I ask a couple of simple Q's?

If you took a bar of White LED's, put in some kind of reflector, could you ever hope to produce enough light to grow plants?

What would the rough cost be for a decent light bar, assuming you're willing to build it yourself?

[dannypd]

The above calculation is base on assumption of a forward voltage of 0.7 Volt, for some LEDs, the forward voltage may differ, I think some are as high as 3-4 Volt.

Amazing calculations!  I will definetaly re-design my layout.  I will also use a stepping (sic??) resistor to allow a "dimmer" effect.

[dannypd]

Ok, so before I brush the cobwebs off my electronics knowledge, can I ask a couple of simple Q's?

If you took a bar of White LED's, put in some kind of reflector, could you ever hope to produce enough light to grow plants?

What would the rough cost be for a decent light bar, assuming you're willing to build it yourself?

I doubt these LEDs would be able to grow plants.  They are just possibly not bright enough, unless you REALLY put them in a HUGE array!!  Then you WILL have heating issues!!  (hot LEDs = sad LEDs)....