[sdiy] question - ripple tolerance

[Harry Bissell Jr]

Hello Merk (et al)

There is no hard and fast rule for ripple rejection.
The question becomes... what is the ripple rejection
required by the circuit.  

Many opamps themselves have excellent ripple
rejection... far in excess of 60dB. This is to say
that
AC signals on the power supply pins are not ampplified
but rather, cancelled out.  Or that the device is
simply not sensitive to supply voltage variation
directly.

The biggest offender in cases of poor ripple rejection
is using the supply voltage as a voltage reference.

Often this is a bias voltage etc... that should be
supplied from at least a well bypassed (capacitivly)
point... or better still from a voltage reference IC
or zener diode shunt regulator.

The methods for ripple reduction in the supply are
usually using more filter capacitance... and adding
output filtering (RC or LC) with a cutoff frequency
far below the 60Hz (ok euros, FIFTY HERTZ :^).

Full wave rectified supplies have twice the ripple
frequency (100-120Hz) so they are easier to filter.

Shunt regulators usually outperform series regulators
in ripple rejection... if you can stand the power
wasted

H^) harry



--- mark s <djarum11 at mindspring.com> wrote:

> question-
> 
> are there desired pre-regulator and post-regulator
> ripple voltages maximums 
> (or minimums..) that are preferred when designing a
> high quality audio 
> power supply ?
> 
> i am under the impression that 1V pre-reg and 1mV
> post-reg (the reg here 
> has a ripple reduction of 60dB) are sort of the
> norm.  is this on the mark, 
> or what is a better value to shoot for?
> 
> mark
> 
>