In article , Iain M Churches
wrote:


"Jim Lesurf" wrote in message
...


I have found that amplifiers with the most pleasing (this may not be the
most accurate:-) sound seem to have a harmonic distortion decreasing
gradually as the order increases.

That may be your experience. However my personal interest is not actually
in amplifiers with a 'pleasing sound'. I don't want the amp to have any
noticable distortion signature. I'm trying to listen to the music without
it being affected by an amplifier 'sound'.


This is perhaps one of the problems with NFB. It cancels the "benign"
even harmonics, leaves exposed the more unpleasant odd harmonics.

I've seen comments like that made many times. However feedback should
reduce nonlinearity. Not sure why NFB should be assumed to act more on
'even' than 'odd' harmonics.

Again, TBH, I am more concerned with intermodulation or distortions that
create anharmonic components as these have seemed to me to be much more
noticable at a given level than simple harmonic distortion. However to get
low IMD tends to imply we'd get low harmonic distortion as well...


However my personal preference is for amplifiers that have minimal
levels of distortions at whever harmonics fall in band.

I have listen extensively to large Crown (Amcron) and Carver broadcast
amps. They have incrediblt low distortion figs which as a builder of
valve amps I can only dream of. They may be incredibly accurate, but
their sound leaves me cold.

Can't comment as I have no idea why you would feel that from what you say.
Hence I could only make guesses.

My main interest is in high quality valve amplifiers, which IMO are
much more of a challenge for a designer than SS (I also happen to
like the way valve amplifier can sound)

Depends what you mean. :-)

I am fortunate enough to be a recording engineer by profession, so I
have for many years had the opprtunity on classical sessions to sit in
the studio during rehearsals and hear the "real" sound, which can then
be used as a reference for what is coming out of the loudspeakers. Even
in this digital age of hard disc multitrack systems, I still use a
Radford STA100 .(UK built 1964) for monitoring.

I was fortunate enough to be employed as an audio designer for some years,
which made me feel like a "kid in a sweetshop" at the time. :-)

My recollection is that Radford made some good valve amps. In particular,
they paid an unusual amount of attention to measuring, analysing, and
designing around the area of phase control, transformer behaviour, and
ensuring good stabilty margins. Alas, my impression is that since then many
of the 'home brew' valve amps people make do not benefit from this.


Yes. The difficulty, though, is ensuring you make the relevant
measurements and can interpret the results appropriately. One of my
niggles with audio reviews is that they often fail to do this and just
give simple standard values that don't tell us much.

I agree entirely. It may also be that on the test bench we put too much
importance on some measurements and do not undestand the relevance of
others. In valve amplifiers for example damping factor has an enormous
bearing on the way the amplifier sounds. One takes a high DF for
granted in an SS amp. In a valve amplifier, a figure above twenty
(which seems adequate in listening tests) is not easy to achieve.

I would certainly agree with the above. This is one of the reasons I am
less than happy with many of the reviews in magazines. People have known
for decades that the output impedance can affect the performance in audible
ways - and that this will vary with the speaker as well. Hence to omit any
consideration of this from valve amp reviews strikes me as 'uninformative'.

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