shinji2k

What I think Ion is trying to say is that the LEDs aren't going to be exactly the same. Some will have less resistance and thus draw slightly more current than others. With all of them in parallel, the current will travel the path of least resistance and eventually burn out those LEDs. LEDs in series will not have that problem. Multiple LEDs in series will behave like one larger LED, each using only the current they need.

I could be completely wrong though .

__________________
Pearl in the Rough (My on going case mod comp project that I couldn't finish on time)
Actual Power Consumption of a few components I own

Please forgive me, I have SIWOTI syndrome.

A still more glorious dawn awaits. Not a sunrise, but a galaxy rise. A morning filled with 400 billion suns. The rising of the Milky Way.

Ionimplant

Here is the simplified equation for a diode.


ID is the diode current
IS is the saturation current
VD is the voltage across the diode
n is a factor based on technology
Vt is the thermal voltage (kT/q where k is boltzman, T temp, and q elec. chg)

If you do what I call the 'bad' circuit will 50 diodes in parallel, they all have the same VD. Guess what! They do not all have the same n. Nor do they all have the same temp (though they are close at first). These terms are the exponent of e. What happens if one diode has an n=3 and another has an n=3.1. (parasitic resistance is not modeled in this eq. but it is different too as well as IS--everything is different!!)

So what happens is the diode with the lowest value of n starts sucking most of the current while the others happily allow it. Then that diode pops and the next one takes the load.

Maybe you would get lucky but it is poor circuit design.

If you really want to minimize the component count, just start with a bigger voltage. Put a dozen AAs in series and build ten stacks of 5 LEDs in series and 5 resistors.

Take a look at this article: http://focus.ti.com/lit/an/slyt084/slyt084.pdf

There are a zillion other articles out there talking about driving LEDs.

I hope this helps.

Ionimplant

It would be "slightly" if it were a linear equation. It is an exponential equation. Big difference.

EntropyTTU

Yeah, but would it not be any different than multiple loads, drawing multple currents. Current does, travel different paths, but its not like it is all going to flow to point of least resistance, only in a short circuit will all of it flow to least resistance.

The parallel idea, would be not different from a parallel circuit with, a 5, 10, 15 Ohm load, they each will pull different currents, but no more, and no less.

__________________

shinji2k

Hey, I was close. But now that you mention it even small variations would have major effects, much more than slight.

__________________
Pearl in the Rough (My on going case mod comp project that I couldn't finish on time)
Actual Power Consumption of a few components I own

Please forgive me, I have SIWOTI syndrome.

A still more glorious dawn awaits. Not a sunrise, but a galaxy rise. A morning filled with 400 billion suns. The rising of the Milky Way.

Boris4ka

It's around 6"x1.5"x1". A standard RC battery. - http://www.microracers24.com/media/VRNVEN-1532.jpg

I can't find a 4W resistor. Can I use four 1W resistors? And cost is a big factor for me too, I don't want to spend $20 on a resistor. If it comes to that I'd rather solder 25 resistors and have it take an hour more.



Wouldn't the difference be negligible? I've never heard of this before. Then how come all LED's aren't wired in series?

EDIT: I posted this before I saw the previous 4 posts. You're saying my lights will burn out? It still doesn't seem right to me, I understand it having a difference with like a bunch of 500lumen LED's but with these?

__________________
[CS:S Gun Game (ADMIN)] [My CPU-Z]
[FAQ - How to make a portable USB charger]
[OCN Wallpapers] [OCN Wallpapers 2] [OCN Wallpapers 3]

shinji2k

Gah, it's been too long since physics class in college. I am of no help here . Things in parallel only have the same voltage, not amperage. Or something. I should go.

__________________
Pearl in the Rough (My on going case mod comp project that I couldn't finish on time)
Actual Power Consumption of a few components I own

Please forgive me, I have SIWOTI syndrome.

A still more glorious dawn awaits. Not a sunrise, but a galaxy rise. A morning filled with 400 billion suns. The rising of the Milky Way.

EntropyTTU

Yes, the circuit in the top picture, is what I proposing to begin with, and it will defintely work. And with the diode equation, I see your point, and you are exactly right, there may be a difference in lifespan of the leds, but I wonder how much difference it will make?


Boris:

Either way, the original picture I posted with each led getting its own resistor is guaranteed to work, if you want to be safe.

But, I think a shot at the 50 in parallel, or 25x2, is worth it too, if that is what you want to try.

Best of luck to you.

__________________

Ionimplant

There are multiple questions here...

"Then how come all LED's aren't wired in series?"

Well...they ARE!


You can use four 1W resistors in parallel (to make the 'bad' circuit)

"I've never heard of this before." If you are an EE in the circuits area, you should have heard it before.

To Entropy's question about multiple loads...multiple LINEAR loads are fine.
Don't ignore the diode equation...it tells all. Look at the TI document.

Reista

Im not going to post anything about the actual array, the guys seem to have it covered from what i read but you might like to look into a one-way motor that will generate a current when you pedal your bike to recharge your battery (if your using a rechargeable.)

__________________

My case: Black Wind
My Bench: Benchy

And remember kids, fans don't Cool the air, they Move the air.
"Official National PEBKAC Association member."
"If my computer had a virus like AIDS, id take it to a LAN party =]"