[sdiy] Re:

[Brock Russell]

> > How is this better and more efficient than a voltage follower and a
> > current-limiting resistor?
>
>This way you'd get linear control over LED light intensity with varying
>control voltage. Remember, as with any diode, a LED doesn't exhibit a linear
>voltage vs. current relationship. In other words, linearly varying the LED
>supply voltage (eg with a voltage follower)  won't produce a very linear
>change in light intensity.

LED colour also changes with current so you have two non-linear
characteristics that will vary with voltage. PWM is the way to go.

Brock