[Scott Dorsey]

Les Cargill wrote:
(Scott Dorsey) wrote:
Les Cargill wrote:
John Williamson wrote:
snip
He/his prefer to select parts at high precision and use simpler
designs. But since this doesn't scale, it'll never be a
mass-market idea.

I think this is a good plan, BUT by the same token I think that it can be
used along with moderate feedback. Note that one of the thing Nelson Pass is
really good at is minimizing the number of devices that require high precision
in a design.


This makes sense, but a lot of his writing is about part selection - not
part *number* selection but selection of specific parts from bins.

This is the downside of trying to use current-technology semiconductors,
especially FETs, without feedback. Pick two MOSFETs off the line and the
curves will be wildly different. You can hand select or you can pay
On Semi to do the selection for you.

Feedback lets you get away with different curves and different gains and
still get reasonable performnce without individual tweaking of circuit
parameters with hand-selection and trimmers.

It's not really true, because the whole push-pull thing nearly eliminates
even harmonics... so all you have left is that which is due to the output
transformer, and classic tube amps were almost all push-pull designs for
that reason.


But there's a fetish for "single ended" as well.

That is a very modern thing. In the fifties, sixties, and seventies,
single-ended output stages were relegated to table radios and classroom
phonographs, but then the Japanese went berserk over the idea in the nineties
and now there are a lot of people selling very highly colored and less than
sonically optimal products at high costs. Go figure.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."