Power Supply Filter Capacitor Size

[Grant Richter]

> >
> >Regulator power dissipation (read heat) is reduced by matching
> >the capacitor to the load. The regulator does not have to dissipate
> >the constant excess input voltage and as log as the bottom
> >of the "ripple" is above the dropout voltage, it should show the
> >power supply ripple rejection the regulator is rated for.
> >
> .....can this be true?
> The dissipation in the regulator is the drop across the regulator,
> multiplied by the current through it.
> I mean, this is Ohms law in practice.

In a DC sense, yes. But considering AC the 50 or 60 Hertz
full wave rectified signal actually forms a sawtooth wave
at the power supply filter capacitor. The peaks of the
sawtooth are the maximum line excursions which charges
the filter cap. Then the circuit runs on the discharging
capacitor, until recharged by the next line cycle.

This waveform has a average DC value with a sawtooth
imposed upon it. If the bottom of the sawtooth is above
the regulator dropout voltage but no higher than needed,
it will reduce dissipation.

This depends on transformer output voltage, desired
under voltage surplus, load range and a host of other
fine points of no interest to the hobbiest musician.

Like Paul says, just make the cap as big as you can <:^)