In article , Don Pearce
wrote:
On Tue, 15 Jul 2003 08:42:20 +0100, Jim Lesurf
wrote:
[snip loads to avoid repetition]

I appreciate all this reasoning, Jim, but the way I see it is this:

ALL of the AC signal current that goes through the speakers must pass
through these electrolytics as series elements.

Agreed. ;-)

That puts them in the signal path.

This is what I find unsatisfactory as a description for the reasons I
outlined. - i.e. the amp should be drawing the current, but rejecting or
ignoring any resulting voltage variations of the rail voltage from which it
is drawing the current. Thus provided the available current and voltage are
'sufficient', the rail variations should have little or no effect upon the
output signal pattern.

In effect, the rail rejection of the amp seeks to isolate imperfections of
the reservoir caps from the 'signal path' in terms of their having any
effect upon the output. This is not so for the other cap locations I
mentioned. Hence the distinction I make.

There *are* (or can be) problems in these areas, of course. For example,
where some form of 'common mode' impedance due to shared power
arrangements in a stereo/multichannel amp causes distorted cross-talk. The
problem with things like this, though, is that unless we take care we end
up regarding *everything* as being in the signal path. It is possible to
argue this, but my worry is that the term may end up being so broad as to
cease to have any value.

My concern here is perhaps analogous to the way some people tend to use
'distortion' to mean *any* change in waveform shape (e.g. due to linear
filtering) whereas others limit it to effects due to non-linearities.
Problems then arise when we try to use the term to discuss specific issues
as the term ends up meaning different things to different people and they
may not notice they are arguing/discussing at cross-puposes as they are
making different assumptions as to the meanings of the terms. As an
academic I tend to prefer fairly limited and specific definitions where
possible to avoid the risk of this.

Whether they are at the "cold" or "hot" end of the speaker is really
neither here nor there.

Agreed. :-) Not the point I was making, though.

Actually my use of the word "ALL" above may seem contentious, but any
current that doesn't go through the caps must go instead through the
rest of the power supply - including the diodes - where it will be
chopped and modulated by the mains. Another good reason for big, fat,
low ESR electrolytics.

I'd agree that good electrolytics with adequate values, low ESR, etc, are
desirable. However I also feel it is the task of the amp to then reject any
remaining fluctuations on the rail. I's also agree that no amp can be
expected to do this perfectly. But I contrast this with the other cases I
mentioned where the amp can't really do anything at all about any cap
imperfections.

Slainte,

Jim

--
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Audio Misc http://www.st-and.demon.co.uk/AudioMisc/index.html
Armstrong Audio http://www.st-and.demon.co.uk/Audio/armstrong.html
Barbirolli Soc. http://www.st-and.demon.co.uk/JBSoc/JBSoc.html