At 04:54 PM 9/18/2007, David Kao wrote...
>That's false.

It's correct. You just don't know what you're talking about.

>By connecting LEDs in serial you force them to all operate at
>a same current but different voltages. This would be a false design if
>all LEDs are different. One LED may be required to operate at 15MA but
>another may be 300MA.

Oops. You either contradicted yourself or are displaying extreme
ignorance.

>If LEDs connected in serial

It's "series."

>  are same LEDs, by connecting them in serial

It's "series."

>you allow them to operate at a same current, which is fine, but still
>each one is forced to operate at a different power level unless every
>single LED is identical.

So? It's not a problem unless you exceed their maximum rating. If you
keep the current to 20 mA, you won't do that (for the _vast_ majority
of LEDs).

>  In reality if the LEDs are not produced from
>a same silicon wafer

Of course they aren't. Most LEDs are made from gallium arsenide. Blue
ones are typically made from indium gallium nitride. I'm not aware of
any which are based on a silicon chemistry (although silicon carbide is
sometimes used as a substrate, due to it's thermal conductivity).

>If somehow one of the LED is shorted out, because of vibration of your
>westy, or just a simple internal failure that usually happens to
>diodes
>as a short first, your other LED will be instantly destroyed too.

LEDs have lifetimes in the 10,000-100,000 range. They cost less than a
buck. So what?