"Arfa Daily" wrote in message
...

"Andy Evans" wrote in message
oups.com...
Note that most transistor amp designs tend to have a flat response, low

distortion, and a low output impedance. Whereas zero-feedback 'SET'
designs
tend to have a tailored response, distortion that rises rapidly with
output, and a relatively high output impedance. All of these factors
may
cause the SET design to change the signal patterns and cause audible
changes in some circumstances.

It may be useful to seperate out some of the commonly used expressions
regarding SETs and valve amps, since I know several have been confused.

a) One of the common characteristics of SETs (though NOT implicit in
the term "single ended triode") is the use of a DHT (directly heated
triode) in the output such as 2a3, 6B4G, 300B and 845 - this can be a
major positive factor in the overall sound, against which can be set
the drawbacks Jim has listed above. .
b) When we come to DHTs we are in another ballgame, in my opinion.
Those who habitually use DHTs, do extended listening tests with DHTs,
and are familiar with their characteristic sound will claim they offer
quite audible gains in clarity over indirectly heated valves. This is
only partly a question of output valves, but in practice only a small
handful of people will have used small signal DHTs for preamps, input
and driver stages of amplifiers. Reason being that while output DHTs
are available in current production, there are zero small signal DHTs
(except one or two rare boutique ones at huge prices). They are
available on ebay since they were used in countless old radios from the
20s and 30s, and there are good but dwindling stocks still to be had.
Some on the other hand have become quite rare. As you might guess, I'm
a huge fan and use them in preference to anything else.
c) DHTs and zero feedback can be used in push-pull amps with different
results from their use in SETs. Each of these factors contribute -
feedback, push-pull/single ended, indirectly heated or directly heated
and so forth (we could go on)

I don't personally use SETs and never have, and it is useful in terms
of general attribution theory to consider what we attribute to which
factors. In particular DHTs and SETs are absolutely not equivalents -
one refers to valve construction and the other to circuit topology. One
might consider that SETs with DHTs are a bit of a curates egg - both
good and bad factors. I'd personally opt for all-DHT push pull amps
given the choice. We haven't even come to another common SET feature -
interstage transformers - but that's enough to be getting on with. Andy

What is the theory behind an indirectly heated valve, having an inferior
performance to a directly heated one ? Not being contentious - just
interested ...

Arfa



Ah. OK ... Looks like I have might have my answer from some of the other
posts that have appeared since I posted my question. Once again, it seems to
be one of those " in listening tests 99% of people could hear a difference "
situations, with no actual published measured data to back the claims up. It
seems to me anyway, that unlike in the ongoing interconnect debate, where
everything stays fundamentally the same bar the cable, in this case, it is
apples and oranges both being classified as fruit ...

Since you can't buy the same valve in directly and indirectly heated
versions, you can't cross substitute to see if one performs differently from
the other in the same amp. Even if you could get both versions of the same
valve, you would have to have differently designed support circuitry for a
cathode-less valve. As Andy has noted, because the cathode now is the
filament, radically different heater circuitry has to be employed.

As far as I understand valve theory and historic development, the only
reason that valves were originally directly heated, is because they were
basically modified light bulbs. Once the makers had got the hang of
producing harder vacuums, it was then possible to use thermo-emissive
chemicals in the valve, which allowed for operation at a much lower
temperature. To improve the efficiency of the emission process, these
chemicals were deposited evenly on a cylinder, heated from the inside by the
filament, threaded up the middle. This also allowed the filament or heater
circuitry, to be separated from the cathode circuitry, making possible
simple AC powering of the heater. All very simple and logical historic
progression, as production technology and materials progressed.

Whether the electrons are forced off the surface of a tungsten filament by
viciously heating it, or produced as a result of applying more gentle heat
to chemicals which readily give up large quantities of electrons, is neither
here nor there, so I can't basically think of any reason that - setting
aside differing electrical characteristics that all valves have designed-in,
dependng on the job they are intended to do - there should or could be any
audio-affecting performance difference between correctly operated valves of
either type.

Perhaps someone who purports to know the difference could enlighten me, as I
really am interested in knowing.

Arfa