On Sun, 17 Jan 2010 22:55:31 -0500, Charlie Smith
wrote:

Unless somebody has any modifications to suggest to V2, then the board
will be built as submitted.

---
I suggest you ditch the PWM and use either Jim's circuit or mine in
order to get the variable dimming that you want.

Why?

Because the PWM circuit, as JL stated is doing nothing.

In truth, it's worse than that because no matter how you slice it the
PWM circuit will generate EMI, which is potential trouble that goes away
completely with a linear dimmer.

In addition, you save no power using a PWM dimmer.

Consider:

If you're using a 14V supply to light up 20 LEDs to full brightness with
20mA going through each of them, then the total power dissipated by
either a linear or a switched supply system running at 100% duty cycle
will be: 14V * 0.4A * 1s = 5.6 watts.

Now, (using a switched supply and assuming a 50% duty cycle will get us
1/2 brightness) if we drive the LEDs with 20mA for half the time, the
system will dissipate: 14V * 0.4A * 0.5s = 2.8 watts.

In the same vein, if we assume that half the current through the LEDs,
all the time, will give us half brightness, then for the linear supply
the dissipation will be: 14V * 0.2A * 1s = 2.8 watts.

If you're interested in blowing off the PWM stuff, I'd be happy to
rework my design to provide input transient protection and supply you
with a schematic, a BOM, and a simulation.


JF