power supply noise

[media at mail1.nai.net]

At 12:49 PM +0100 02/05/98, Haible Juergen wrote:
>I never made serious experiments on this, but reportedly it's not that
>important
>to filter out everything down to a couple of hertz if you are using
>*opamp*
>circuits.
>Opamps tend to have excellent power supply rejection at low frequencies,
>but poor rejection as frequency increases. So maybe even combinations
>of 22 Ohm / 10uF (or something similar) make sense.

Actually, I've swapped out 741's for newer IC's in audio circuits and have
heard a noticable decrease in hum at the output (what is the PSRR of a 741,
anyway??).  Newer op-amps also produce less thermal noise.

It is true that power supply rejection decreases with frequency, but power
supply noise also decreases with frequency when using bypass caps.

I've used both 10uF and 1uF (or 1uF and .1uF) shunted to ground with a
resistor between them, but as you say, this is probably only useful for
discrete circuits.

So why do many of you say a 7815/7915 are not good enough a modular's PSU??
Noise??  Stability??

I vaguely remember a circuit using a 78xx regulator with both a cap at the
output and additional circuitry between the ground pin and ground to reduce
ripple.  However I can't remember what it was.  It wasn't a diode to lift
the output of the fixed regulator to some higher value and it wasn't
anything to increase to current output.  Any ideas??

PEACE OUT :)
MARK