Do you have a link to the Aeroelectric forum that you are refering to?
If you don't think the battery acts a filter, you've likely never seen the output of an alternator on an oscilloscope.
Below is a March 10 2013 quote from Bob Nuckolls on the Aeroelectric List that I found for your consideration. Would be interested to hear reactions; did I overstate the case in my previous post?
regards,
erich
"Later on I offered a revision to the ol' mechanic's
tale that a battery was one of the 'best filters of
noise' on the bus. I used to subscribe to that
notion myself. My colleagues believed it.
But consider that a battery has three states of
existence in the properly functioning system.
We KNOW that an engine driven power source must
be adjusted for at LEAST 13.8 volts in order to
fully charge a battery at 20C. Setting it somewhat
higher, say 14.2 to 14.6 overs a faster replenishment
of charge after cranking the engine and perhaps
a prolonged battery-only pre-flight activity.
At the same time, we know that turning the
alternator off allows system voltage to fall
quickly to 12.5v or so . . . it marches downward
from there with time.
So what happens in the range of 12.5 to 13.8
volts? Not much. The battery is incapable of
taking on significant energy below 13.5 and
doesn't start delivering significant energy
at more than 12.5 volts. I.e. the battery becomes
'unhooked'. The idea that it takes on the role
of some 'super capacitor' for the smoothing
of noise is without foundation. I've verified
this both on cars and airplanes where disconnection
of the battery while the alternator is running
produces a only a small rise in bus noise.
A 14v alternator is a low impedance source with
about 1.5v pk-pk ripple built in as an artifact
of 3-phase rectification. This is a given that
drives DO-160/Mil-STD-704 requirements that
qualified devices be designed to function
as advertised in the presence of such noise.
The noise spectrum to be tolerated is plotted
here . . .
http://tinyurl.com/b3rhjwq
This is a 28v system plot where worst case
noise is 1Vrms over the range of 1 to 5KHz
and falls off on each side. 1Vrms sine wave
is 2.8Vpk-pk. Any departure from sine wave
allows the pk-pk values to rise markedly. In
short, the battery is not an effective filter
for anything except gross brown-out events and
standing off a runaway alternator for tens
of milliseconds required for the ov protection
system to assert system shutdown.
This ol' dog didn't learn it until about half
way through a 40+ year career. Not that it
was hard to figure out . . . just didn't have
a reason to question what was proffered as
common sense."