Scott wrote:
brian whatcott wrote:


Quite so. Moreover, the idea of providing an individual series
resistor to pass about 20 mA works for a good variety of LEDs.
Focussing on the desire to dim a LED, there is no reason (that I
know) not to reduce the current. This has the great advantage over
pulse width modulation that it does not kick up radio interference.
There are LED drive chips on offer now that might suit....

Brian W

Maybe this will help the builder. A current source (rather than a
voltage source) will make the LED brightness varaiable. A typical (red)
LED runs at about 20 mA for full brightness, while dropping about 1.7V
across the LED. Lowering this current dims the LED. The shown circuit
can be set to supply up to about 400 mA, enough for about 20 LEDs. Of
course, he will want to use red LEDs in place of the white ones drawn in
the circuit...

http://radiolocation.tripod.com/LEDd...ampDimmer.html

Scott


Yep.

Here's another idea that sticks with the non pulsed approach.

It uses a variable voltage reg in constant current form - allowing a
string of several LEDS as is:
http://www.instructables.com/id/Circuits-for-using-High-Power-LED_s/step11/an-even-simpler-current-source/

...or in TinyURL format:
http://tinyurl.com/yb9kyyr


....this could be massaged to provide the drive for 24 LEDs at rather low
cost and component count.

Brian W