Walther said:
I understand that point. But as I found out, back in 1975 there
was a big contest in a british hifi magazine for expensive amps.
Both the amps and the testers were from the who-is-who of the
high end hifi amp scene, but the magazine covertly added some
candidates from an earlier epoch - one of which was the STA-25.
To everyone´s big surprise it clearly won that competition.
Doesn't surprise me.
But don't forget that there are mainly 2 ways of thinking in amplifier
land:
#1. The "straight-wire-with-gain" adepts (objective thinking) , of
which Radford and Quad are fine examples, and
#2. Those looking for "pleasant' tube sound, whatever that may be.
That's entirely subjective thinking.
It mostly boils down to a certain amount of 2nd order harmonics, as to
be expected when using SET amps or unbalanced PP amps.
Back in '75, the race for ever lower distortion figures (THD
amplitude, not harmonic spectrum) was going on.
Remember Luxman with 0.0001% THD?
The reason why I'm modifying or even rebuilding old amplifiers, or
even creating new designs with old trannies, is that I'm not a Walker
follower. I respect the POV of Walker, Radford et al, but one must
realize that within the technical and commercial boundaries of their
time, they just were looking for low distortion only.
These days, most tube users acknowledge that there's more to this
subkect than just distortion and high NFB rates.
The spectrum of the generated harmonics is way more important than
just the amplitude of harmonics.
With this, other factors come into play, like IMD, side effects to
global NFB, bandwidth, power supply design etc.
These days, components are of far better quality than ever.
It is most certain that building a "new" tube amplifier of ancient
design with modern components, perhaps adding solid state circuitry
for protection and supply stabilization, will lead to an amplifier
that measures better and sounds better than the original.
We're in a position that not one of the famous designers was before,
that we have a history of designing where we can choose whatever
circuittry or components to reach our goal.
Imagine that Walker or Radford could have used modern day solid state,
do you think they'd bother with glassware anymore?
Usually, when an old amp falls into my hands, and it's not a
collectable, I strip it entriley, and mostly only the trannies are
used for new projects.
Good idea, but I´ll better leave my STA-25 untouched.
I'd probably do the same, because a radford STA25 is a collectible,
and will keep its value over time.
However, there are dozens of other amps from the '60s , like the
Lafayette, Bocama, Wenbley, Sansui types that would benefit enormously
from a comlete overhaul.
The transformers in said amps are sometimes of very good quality,
something that the old design couldn't ref;ect in any way.
Example: Philips HF309. EL84 PP monoblocks with ECC83 phase splitter.
The transformers, AD9058 UL tapped, have a bandwidth of over 100 kHz.
The Philips amps barely cover 25Hz...30 kHz.
When used with say a pair of 2A3 in PP, a very good and still low cost
PP triode amp can be had that puts out 10...15 watt (comparable to
EL84). Without global NFB, bandwidth may extend to 80 kHz -3dB.
The 2A3 puts out some 2nd harmonics, but very few 3rd and other odd
harmonics. Because of the PP topology, even order harminics are
cancelled out when the tubes are properly matched.
So, we have about 0.3 % THD up to half power *without* any NFB added.
When NFB is applied, the figure goes down even further.
One of the main advantages of zero NFB is that the distortion rises
slowly with increasing power, and some headroom is present for peak
signals.
A penthode amplifier with heavy doses of NFB, like the Radford, will
have lower THD, but reaches a brick wall at max. power.
This is something that Peter Walker did netter IMHO: the power reserve
of a Quad II is about 40 watts, while it is just specified as a 15
watt amp (in 16 ohms!).
The Quad II is in fact a class B design, but due to the specific
output ransformer circuit (cathode feedback windings, a form lof
*local* feedback) without the particular class B disadvantages.
McIntosh amps sport a similar transformer.
Nowadays, Audio Research follows an even simpler route by using the
secondary winding as a cathode feedback path.
Instead I´d like to build me a copy of it with state-of-the-art
components... and I of course would like to know if there has been
significant progress in tube valve amp circuitry design since 1975!
Tube amp circuit design stalled about 1965, perhaps even earlier.
Quad and Radford stopped making tube amps in the '70s, but the designs
are from the '50s.
Only the Japanese never forgot about tubes, and e.g. Technics designed
a fine OTL amplifier in the late '60s with special tubes 50C5 or
something.
Very rare, very expensive.
There are many newer ways to improve amplifier performance since then.
The availability of solid state may be of help as well.
Hybrid amps is another example of a topology that didn't exist until
1985 or so.
Earlier I had the idea of an improved CSV-60 clone which has been
suggested somewhere on the net - but now the STA-25 looks like a
better platform to me.
The phase splitter is probably one of the first things to look at.
IMO, the floating paraphase circuit sounds better than the ECF82
circuit of the Radford.
SETAs is an entirely different subject, I mostly build PP amps.
I must confess I´m a total rookie... I thought of doing a layout
for a STA-25 clone PCB, add some ground layers and shielding.
I'm not a fan of PCBs in tube amps.
Not simple to change something, and it is said that PCBs destroy the
sound. Something I haven't heard, but people like Menno vanderVeen did
research about that and found out that point to point wiring is almost
always the better option.
Maybe I´d better keep this for myself until I got it working...
Want to suggest improvements? Go ahead: walther at in4tec dot de.
I'd figure we'd keep this in the group, so that others can join in and
we all might learn something! :-)
--
Sander de Waal
" SOA of a KT88? Sufficient. "
|