mick wrote:

On Tue, 16 Nov 2004 00:28:23 +0000, Doug Schultz wrote:

I dont see why anyone is giving Andre a hard time here. this is really the
first decent project post that I have read in RAT. I couldnt possibly
afford to build something like this right now. I am stuck with using the
stock of Russian 6L6's as triodes for now. And whatever I can find in my
Junk box.


I'll second all that - especially about the stock of russian 6L6s as
triodes! That's just what I'm listening to, and pondering about whether to
get a couple of 2A3s - they are about 22 quid each instead of 80 quid
each for 300Bs! Both of them make the stock 6L6s (I've got 4 left) look
good when my wallet is this thin...

Actually, as this amp uses the russian 6L6s marked 6L6C (that are
really 6P3S and more like 6L6B) I might just build a new 300v PSU,
parallel 2 pairs and get some new OPTs for it. Now, should I SE them or
p-p them in triode mode? At the moment I'm leaning toward SE with fixed
bias for about 3W from a pair. Don't you love power? ;-)

Trioded 6L6 can give 6 watts each in SE triode class A, so a quad of them
will make about 24 watts, same as a 211, and the whole thing will be less
lethal,
and easier to make.

Poor man's audio nivirna is possible with low voltage tubes.



Patrick Turner



--
Mick
(no M$ software on here... :-) )
Web: http://www.nascom.info
Web: http://projectedsound.tk

On Thu, 18 Nov 2004 22:46:18 +1100, Patrick Turner wrote:

snip

Trioded 6L6 can give 6 watts each in SE triode class A, so a quad of them
will make about 24 watts, same as a 211, and the whole thing will be less
lethal,
and easier to make.


grin Thanks.
Hmm... RCA list 6L6/6L6G at at 5% THD for 1.4W. I bet 6W clobbers that a
bit! I see that you can push 10W out in boosted triode mode. That would be
entertaining to try... I haven't tried anything other than spec sheet
parameters so far. I should be able to give these (63PS) another 50v
before they reach the limit for g2, but I doubt if it will add that much
power output.

--
Mick
(no M$ software on here... :-) )
Web: http://www.nascom.info
Web: http://projectedsound.tk

mick wrote:

On Thu, 18 Nov 2004 22:46:18 +1100, Patrick Turner wrote:

snip

Trioded 6L6 can give 6 watts each in SE triode class A, so a quad of them
will make about 24 watts, same as a 211, and the whole thing will be less
lethal,
and easier to make.


grin Thanks.
Hmm... RCA list 6L6/6L6G at at 5% THD for 1.4W. I bet 6W clobbers that a
bit! I see that you can push 10W out in boosted triode mode. That would be
entertaining to try... I haven't tried anything other than spec sheet
parameters so far. I should be able to give these (63PS) another 50v
before they reach the limit for g2, but I doubt if it will add that much
power output.

The 1.4 watts is for very tame operating conditions.

But with 400v at anode, and 50 mA of idle current,
that's 20 watts, and if efficiency is 25%, you get 5 watts at least.

Patrick Turner.



--
Mick
(no M$ software on here... :-) )
Web: http://www.nascom.info
Web: http://projectedsound.tk

On Fri, 19 Nov 2004 19:18:52 +1100, Patrick Turner wrote:

snip

The 1.4 watts is for very tame operating conditions.

But with 400v at anode, and 50 mA of idle current, that's 20 watts, and if
efficiency is 25%, you get 5 watts at least.


Good point! I'll knock up another test rig and see what happens.
I've found the data on these valves now, if anyone is interested (the spec
looks a lot like 6L6B):-
http://www.tubes.ru/techinfo/HiFiAudio/6p3s.html

--
Mick
(no M$ software on here... :-) )
Web: http://www.nascom.info
Web: http://projectedsound.tk


--
Mick
(no M$ software on here... :-) )
Web: http://www.nascom.info
Web: http://projectedsound.tk

KISS 101 by Andre Jute

Fabio:

I've held this reply which I wrote last week because we're getting
ahead of ourselves here.

This isn't in the first instance an engineering project. It is an
exercise in hedonism with engineering assistance. Maximum power is not
a consideration as you can now see in the latest installments.

However, there is something frightening about your -95V bias for 300B
with which we should deal immediately lest someone is misled. Here are
some numbers for a high voltage, high current 300B design I made
several years ago that I and others have built and found most pleasing
over a period of years. Zero signal operating voltage 385V (with
cathode bias that means an anode voltage of 465V, eh? - - read the
tube spec again!), bias -80V, quiescent current 80mV. OPT primary
impedance 5K6. That is a design that runs around 80 per cent of
maximum dissipation, which is as high as you can go without utterly
deserting conservative design. Go up to -85V, 85mA (never mind your
-95V and 85mA!) and chances are you'll always be worrying that you are
chewing up that rare/expensive tube for no good purpose. Even down at
the 32W dissipation level, you don't do it for the power but to put a
very high impedance on the plate to flatten the load line and thereby
to reduce distortion and, most importantly, to redistribute the
harmonics more pleasingly. Those matters are described in sections I
have since published. The current (KISS) design is much more
conservative because in my own version I will use genuine WE tubes,
which I am no hurry to take beyond conservative dissipation as I want
them to last the rest of my life, and because this is really not a
series for the experts on RAT but for beginners in whose amps one must
leave a margin for errors and accidents which if the amp were designed
right up to the margin could be catastrophic. The amp you know as KISS
in-house is called The Tradition for a very good reason!

All the classic DHTs I know that are suitable for audio work have low
gain. The only high gain DHTs I know of is in the very recently
designed Svetlana 572-xx range and they are less suitable for (my)
sort of audio work than the more linear tubes with lower gain in the
same range. The low gain DHTs that make higher power produce it by
operating elevated voltages and some, of which the Svetlana SV572-xx
are the most affordable, do it by permitting elevated levels of
current as well.

I don't believe the 417A is that rare, though I imagine those maniacal
collectors in Asia (1) have been driving up the price of the gennie
WE item. Ericson in Sweden made a gold pin version that only a few
years ago people couldn't give away. Check it out under the equivalent
tube number 5842.

The 6S45P is actually a substitute for the WE437 or the similar STC
3A/167A. All three make a super one tube micropower amp. But to use
the Russian tube as a driver you need matched tubes, which is a pain.
If you have a well-matched pair of 6S45P to spare, I will buy them
from you. The WE and the STC are too rare and expensive even to
consider as a sub for the 417A.

Andre Jute

(1) I swapped a mixed batch of five broken and worn out Mullard EL34
plus some Genelec KT66 boxes in poor condition for a matched quad of
Chinese 300B, a matched pair of 845, and carriage from Japan (which is
an expensive consideration because nobody rips you like the Japanese
post office). That deal was made with a dealer, who presumably knew
what he was doing, who actually saw my stuff before he agreed the
price. K-k-k-krazeeee!

"Fabio Berutti" wrote in message ...
AFAIK the problem with a 300B is that it has a very low gain and a
relatively high input capacitance, therefore it requires a driver providing
a high voltage swing and a fairly low output impedance. The 417 used with
an RC coupling will surely perform adequately, but (I'd say) only if full
power is not a goal. From the plate curves, max. power should be obtained
at about 5k anode load, with 450V anode voltage and 80-85 mA, corresponding
to appr. -95V bias, - 190V p-p at least should be needed to get 10W. It
seems a bit too much for a single triode in RC coupling. Surely, with
efficient speakers 5-6W would be enough and the driving voltage will reduce
quite a bit. I thought of other tubes, but after all the 300B is easy to
find and not too expensive (plain Sovteks or JJ do their job without costing
mega-$); when used conservatively it will last for years.
I don't know how it will sound, but it is a SENSIBLE project: there's so few
that can go wrong in it.
As per the 417 (which is rare here): on this side of the Ocean it is
probably easier to get an EC86, which is not the same tube but not very
different. The Russian 6S45P could be another low-cost solution.




"Anonymous" ha scritto nel messaggio
. ..
KISS 101 by Andre Jute
This text is copyright Andre Jute 2004 and may not be reproduced except in
the thread KISS xxx on rec.audio.tubes
THE VOLTAGES IN THIS AMP WILL KILL YOU. GET EXPERIENCED SUPERVISION IF IT
IS YOUR FIRST TUBE AMP.

INTRODUCTION

This is a project to design an ultra-fi tube amp on the net. The project
is aimed at the rawest newcomers to amps. Anyone can pitch in but keep it
simple, keep it on topic, and leave the personalities off so the threads
do not degenerate into another flame war.

The amp will be a two-stage resistance capacitance (RC) coupled 300B. The
intention is not to build a cheap amp. I shall be designing and building
my copy with parts I have to hand, which were selected for sonic glee
rather than by price. Those who read audiophile newsgroups may consider my
mix of outrageously expensive and good value parts bizarre. Ive written
about classical music since I was a teenager, Im a constant concert goer,
the rest of my audio chain is best quality gear, I have good test and
measuring gear when I want to supplement my ears, I have access to
world-class professional musicians to play live for me as a test against
their disks, and I have designed and built quite a few amps, so when I say
a component is chosen for sonic reasons, I dont mean because it is cheap
or sounds like a boombox.

The amp will be simple enough to be adapted to parts you have on hand or
can afford to order. The idea for those who are absolutely new to the
hobby is to swap in parts until you are happy with the sound and then to
start changing the design of the amp until youre happier still. It is not
my amp, it is yours. Do what you please as long as you take the usual
safety precautions.

GENERAL TOPOLOGY

The official name of this amp is the Real McCoy Type 39 Mk VI Amplifier
The Tradition because it is the 39th hi-fi design I drew and the sixth
major topological iteration or significantly variant design group based on
the original in the ten or twelve years since. Its short name, The
Tradition is obvious as it will be a very traditional design. I have
built this amp many times in different versions, so I know where the
design process will arrive. But we shall take it stepwise all the same, so
that you can design something else with the calculated steps. Im going to
shorthand both the product we design and the process The KISS Amp. KISS
stands for Keep It Simple, Stupid.

Where we shall arrive is at a 417A driver resistance capacitance (CR)
coupled to a 300B power tube which is output transformer (OPT TRX) coupled
to the loudspeaker. The signal section will be sensitive enough to
constitute a linestage taking its signal directly from most CDs without a
preamp. The output will be enough to drive sensitive speakers.

The power supply will be tube rectified, choke input.

Whether such an amp is high-level hi-fi or ultra-fi depends on the quality
of the components selected, and on the layout and construction. It will
have zero negative feedback and will operate strictly in class A1, never
drawing grid current or approaching cutoff under any condition of
operation. It will, if you build it with my recommended components and
values, be utterly silent, with a pleasing harmonic distribution.

GETTING STARTED: WHY WE START DESIGNING AN AMP AT THE SPEAKER

The problem with designing a good audio component is that it requires
thought about matters we are inclined to take for granted. Ask yourself,
What is an audiophile, and what is his relationship to his amplifier?

An audiophile is a man who listens to music. The music comes on some sort
of a disk and is played on some sort of a speaker.

The source is usually the most easily changed component. It is often the
one on which the audiophile has the least choice and over which he
exercises the least control. In the case of the CD, it is also the one on
which for almost any amount of money over a not very high minimum the
return will be disappointing. For the purposes of The KISS Amp we shall
assume the music arrives at the amplifier from either a pre-amp or a CD
player capable of putting out 2Vrms. Most CD players can, and most preamps
put out much more voltage.

The most important element of an entire audio chain is the loudspeakers.
It is the loudspeaker which interacts directly with the listener. If the
loudspeakers are poor, the finest source and amp in the world will not
make the music sound good. We are about to design an expensive amp. My
best advice is that your speakers should cost at least as much as your amp
and possibly more. With the exception of some novelties, every single pair
of my speakers cost at least twice what my most expensive amp cost to
build. Even my commercial amps are all cheaper than my speakers. The
purpose of DIY audio is to let you build hi-fi of caviar quality on a
hamburger budget. The amp we will be designing and building in a shop will
cost you between twelve and fifty thousand American dollars you can easily
build a version of it that will come very near for under a thousand
dollars.

The thoughtful audiophile therefore chooses his speakers first, then his
sources and only then does he build the link, his amp. That is the answer
to my question, a real audiophile has no relationship to his amp unless he
is the sort of person who tells people what his Rolex cost or has a
nodding dog in his car to distinguish it from all the other family sedans.
He relates to his speakers. Practically, you need to know how many watts
your speakers require to drive them before you can decide anything
whatsoever about the amp.

It is generally accepted that the best loudspeakers use point-source,
full-range drivers rather than multiple drivers. Coaxial cones are
point-sources and the Quad electrostatics mimic a point source extremely
effectively.

The speakers we shall design for are horns with 8in drivers of about
100dB. Specifically, we are designing for Lowther PM6A fitted to Fidelio
type bicor cabinets. You can buy drivers and cabinet-wood and build them
yourself for a huge saving off the ten grand plus dollar price of a pair
delivered at your door. I built mine. The amp will also drive horns with
similar Fostex and other drivers, and of course smaller horns and Voigt
pipes. I shall also publish a high-sensitivity loudspeaker design called
The Impresario (because I created it for a music promoter to use in his
office) which uses a universally available guitar fullrange coaxial driver
and so can be very inexpensively built. (Yeah, youre absolutely right.
There is no such thing as a free lunch. It will turn out to be a big
speaker even if economical and easy to build.)

Finally, by simply selecting a different output transformer (OPT TRX) or
wiring up the one I recommend differently, the amplifier will be suitable
to drive speakers of a sensitivity of 93-96dB.

Single 300B are not suitable to driving speakers under 90db sensitivity
though I have used a pair in parallel single-ended (PSE) to drive Quad
ESL-63 which are around 85dB sensitive with surprising results.

NEXT PART:
How much power will our amp require?

/5