[sdiy] voltage and LEDs

[db]

So do these LEDs actually need 1000ma to drive them, and a single 7805 is 
pushed to it's limit but powering one?
So maybe a 2ohm 5watt resistor in series?

I'm still having a hard time with this concept of "constant current".
What makes current constant?
Not just a regular power supply?


At 2/11/2010 07:41 AM, George Mattson wrote:

>The forward voltage is what the diode is going to use up across the PN
>junction. It just "goes away"
>Actually, that's the voltage required to get the current to cross the
>barrier region. Consider it the PN road bump.
>
>Using 5 V, subtract the 3.9V Vf from 5V to get the difference of 1.1V.
>1000mA max current is 1 Amp. Using ohms law divide 1.1V by 1 Amp to get 1.1
>ohms to limit the current.
>
>That's the minimum resistance necessary to prevent burning up the LED in
>regards to the specs. Anything less will just dim it down. Usually, you can
>get by with half the current. Anything above a certain limit doesn't let the
>LED get any brighter, just dissipates as heat at the junction and shortens
>the life of the semiconductor.
>
>But, 1 Amp is also the high limit of the 7805
>
>You'd either need a huge heat sink and a fan to keep it cool. Or, use two in
>parallel and let them split the load.
>
>The 1.1 ohm resistor at 1 Amp is going to dissipate 1.1 Watts of power
>(P=IE)
>
>I'd suggest using a 2 to 5 W resistor and mounting them above the board to
>let them get some air to cool down.
>
>
>My $0.02
>
>George Mattson
>
>
>I want to drive a 3watt led
># 1000mA - Max Drive Current
># 3.90Vf - Forward Voltage
>
>Can just use a 7805 to get 5v, and then drop it with two diode in series to
>get 3.6v?
>Or do I still need resistors in series with these LEDS?
>
>
>dennis.barton:skylab2000[socal]
>
>
>
>: 9.0.733 / Virus Database: 271.1.1/2680 - Release Date: 02/10/10 11:38:00


dennis.barton:skylab2000[socal]