"Andy Evans" wrote in message
...
I enclose the following, from Geoff Mead (Audiocircle): does anybody know
anything about this? Quote goes:

Re the current thread on speaker cables and sonic differences between
different amp/cable/speaker combinations. There has been an interesting
series
of articles in Electronics World staritng June 2004 if anyone hasnt seen
them.
The author Graham Maynard started looking at the old chestnut as to why
valve
amps sound different from S/S ones particularly when played loud. To cut a
long
story short he proposes that Miller compensation and series output
inductors
are responsible in causing what he calls First Cycle Distortion. His
proposition
is that the reactive impedance characteristics of most loudspeakers
interact
with s/s amps that have these features over the first cycle of each
frequency
component that makes up the music waveform. He suggests that these effects
disappear after the first few cycles and therefore do not show up in pure
sine
wave testing. An interesting conjecture, yet to be fully proven, but the
point
is that
the reactive elements of connecting cables would also figure in this
argument.
If this were true then valve amps should be much less affected ( SE types
excepted?) Geoff Mead

**Nope. There are a goodly number of reasons why valve amps *may* sound
better, when driven loud, or connected to difficult loads. The output
inductor *may* be partially responsible. Here's my thoughts:

* Valve amps Voltage limit rather gracefully. They do so, because not much
Global NFB is normally employed. Clipping distortion tends to contain lower
levels of high order harmonics. The output transformer tends to limit the
production of these harmonics as well. IOW: Valve amps can be pushed harder
than typical SS amps, whilst still remaining somewhat listenable.

* Valve amps current limit rather gracefully. They do so, due to the
horrendously high internal impedances of the tubes themselves. Again, since
Global NFB is usually quite low, gross amounts of distortion is usually
avoided. Typical SS amps employ current limiting systems which sharply
affect the waveform, when triggered. This effect is exacerbated by the
Global NFB system.

* Valve amps demonstrate a partially 'current source' output characteristic,
which, when presented with a rising impedance characteristic (such as the
rising impedance of the bass driver in a typical enclosure) will deliver a
far higher output Voltage and more apparent power. Again, valve amps may
sound louder than their SS equivalents. Ironically, triodes are much closer
to BJTs in this area, since triodes act more like Voltage sources, than
pentodes. I guess that is why triodes are more highly prized.

* When driving VERY LOW impedances, the output inductor in most SS amps, may
present a significant effect on the Voltage output at high frequencies.
(You'll note that very few amplifier manufacturers quote damping factor at
20kHz).

* The output transformer in valve amps, is it's best and worst attribute.
The transformer adds significantly to cost and mass and severely affects
linearity, frequency response and damping factor. On the plus side, the
output transformer effectively isolates any load anomalies from the active
devices.

Of course, none of the above need be a problem. Good SS amps do not require
output inductors, Global NFB, or audibly problematical Voltage and current
limiting. Then, the advantages of SS amps (low distortion, wide power
bandwidth, high damping factor, excellent load tolerance) can be realised,
with none of the downsides normally associated with SS amps.


--
Trevor Wilson
www.rageaudio.com.au