keithr wrote:
>
> "Trevor Wilson" <trevor@[EMAIL PROTECTED]
> wrote in message
> news:67l564F2os07sU1@[EMAIL PROTECTED]
> >
> > "keithr" <keithr@[EMAIL PROTECTED]
> wrote in message
> > news:48155794@[EMAIL PROTECTED]
> >>
> >> "Trevor Wilson" <trevor@[EMAIL PROTECTED]
> wrote in message
> >> news:67k7ilF2onhd7U1@[EMAIL PROTECTED]
> > **The output stage uses no feedback loop. All feedback is local only.
IOW:
> > Degenerative feedback only.
>
> Maybe a little feedback around the output stage would eliminate the need
for
> such close matching without changing the sound, but I suppose that in
audio,
> everybody has to have their own gimmick to differentiate their product.
Indeed the mention of no global FB and local only is a gimmick.
I have the ME 75 amp schematic in front of me right now.
Its the only ME amp schematic I have, aquired before I went online from
goodness knows where;
must have been from Peter Stein during better times because I recall
having to repair an ME
75 with serious faults.
The schematic is an untidy mess of thing, and there are no notes.
Just what you'd expect from a cowboy outfit.
The output stage for a channel has 12 transistors in total, with a type
of Sziclai
arrangement around the 4 or more output transistors.
The Sziclai has two stages, an emitter follower which is arranged with
common emitter output pairs
to create current to ensure the output voltage follows the input voltage
applied to the low
power follower input stage.
Sziclai stage working is all viewable in Google if a search is done.
The input follower does have a low collector resistance and some
inverted gain to work the
output transistor bases, and their collector current action on the
output load creates
the NFB voltage applied to the input transistor emitters.
Lots of open loop gain, and lots of loop NFB are in action.
The input section of the amp has 8 transistors and a damned opamp.
It is has basically a symetrical pair of npn and pnp differential pairs
paralleled so that there are two inputs, one for input signal,
the other for the global loop NFB which Trevor denies exists.
Its global NFB around more than one stage, and Trevor Wilson tells lies
about
a circuit he should know very well.
The output of the two diff pairs power a quad of
npn and pnp class A series complementary transistors to create the drive
to the output
stage.
I use a very similar drive arrangement in my own amps.
http://www.turneraudio.com.au/solidstateamps1mosfets.html
This input topology is fast and has faily low open loop THD of only
about 4% without the global NFB.
The symetrical input diff amp and Voltage Amp Stage was invented a long
long time ago
by someone bright spark way ahead of Stein.
The ME has global FB from the output of the VAS drive amp fed back to
the second input ****t of the parallel
npn and pnp input pairs.
A huge amount of global NFB is used to get the VAS distortion low.
It would be impossible to combine all the many stages in the driver amp
and output
amp of the ME, and have one loop of global NFB. The circuit is too
complex to
ensure stability; ie, it would oscillate badly at HF with so many stages
and only one NFB loop.
In my amp and so many other designed by such luminaries as Douglas Self
and
countless others, the amp is kept much simpler than the ME way,
and only one loop of global NFB is required.
I use mosfets in source follower mode because they are so very easy to
drive,
and its so very easy to make such a circuit unconditionally stable
with ONE good loop of NFB.
People should read ......
http://www.dself.dsl.pipex.com/ampins/ampins.htm
Douglas Self is a guy who knows at least 40dB more than Wilson will ever
know.
The dself website is a good example of what ME should have at their ME
website.
I hope I make myself painfully clear to Wilson and Stein.
> >> After all, even with global feedback, I know of
> >> no design that takes the feedback loop back to the input of the
pre-amp
> >> so it is not entirely global even in that case.
> >
> > **The term 'Global NFB' is usually taken to mean that the feedback
loop
> > operates from output to input (of the power amp stage).
Wrong.
There are usually 3 stages in a power amp, input pair/s,
VAS, and transconductance output stage.
Global NFB is applied around all 3 cascaded stages.
But where you have two stages in an input plus VAS arrangement
with global NFB around two stages, its global NFB, right!!!!
Don't ****ing bull**** to us Wilson!
And where you have two stages in the output stage, and global NFB, its
global NFB.
It should come as no surprise that in fact the NFB applied in ME amps
enables the stages to be stabilised quite well against sprurious RF
oscillations.
Wilson has never been able to explain it properly, as he prefers to sit
high upon
the bull**** bandwagon, and brag about having "zero global NFB"
Having zero global NFB around a coupla stages in SS amps is a complete
curse upon the sound.
SS amps need all the NFB they can muster.
The ME output stage is the one most prone to creating distortions
because it operates
mainly in class B, and the crossover distortions need to be supressed by
NFB action.
To avoid noise from the PSU getting into the output stage
that has barely enough global NFB, ME use a truckload of capacitors to
filter the rails.
Not a bad idea, for it gives the FB network less work to do to clean up
the noise.
The transistor matching also helps, but the FB and Sziclai arrangement
has the majority effect, with matching making only a tiny
contribution to fidelity.
If we consider Quad's 405 "current dumping" amp, the output transistors
are arranged so that it is entirely non critical whether they are
matched or
not, and bias isn't critical either.
Does anyone think ME amps sound better than a Quad 405?
Did Peter Walker have a few more brains than the guys at ME?
Is the Pope a catholic?
I understand the emotional attachment Wilson might have for the ME
design he undertsands so poorly. Its just a silly emotional attachment.
Patrick Turner.
> >
> > Trevor Wilson
> >
|
