Reminded of my roots by Al Marcy (tubegarden) in his recent posts, I
reprint an exchange with someone who built a $400 tube kit and wanted
to improve it:

*******
From: Will )
Subject: what should I do now?
View this article onlyNewsgroups: rec.audio.tubes
Date: 2003-10-20 21:07:19 PST

Ok,

I built a kit (bottlehead) and I love it and Im going to build another
but
not until later (xmas gift, win lotto, etc).

I still feel like I want to know more about what I am building than
connecting terminal 14 to terminal 2. What is a good book to read or
project
to build (cheap) to help me understand audio equip better. Ive taken
multiple electricity classes @ a tech college + factory training level
(Im
a BMW Tech) so I have a good understanding of the majority of most of
the
tube project schematics Ive seen (and general electricity) with the
exception of the actual tube portion itself.

thx,

-will
********

And my reply:


Will

I built a SEX kit from Dan Schmalle (Bottlehead) for a review in Glass
Audio about six years ago.

You dont say which kit you built. But it is worth considering that Dan
originally designed his amps to be of use to those experimenting with
tubes who dont have facilities or want to spend the time making
metalwork (normally the most time consuming part of building an amp
unless youre really tooled up and experienced). The name of his first
amp, Single Ended eXperimenters Amplifier, tells you the purpose.
Unfortunately most people who built it wanted to preserve it lovingly.
General purpose idiots on RAT (there are always a few here) abused me
for putting mine in a large and sturdy but ungainly box with plenty of
space for additional and large components. As Dan intended when he
designed it, I made some pleasing experiments on it, and used it for a
while to break in Lowther drivers, before passing it on to someone
else. Dan soon developed that super little amp into several others. It
may be that your present Electronic Tonalities amp is suitable for
experimenting on.

My GA article demonstrates a few things one can do with even a
well-designed amp to study various changes.

A few important parameters that are not obvious even to someone
knowledgeable about electronics, to be gained by working with actual
amps and your ears as measuring instruments in addition to whatever
else you have (I started with only an old AVO analogue meter), include

adding chokes to the power supply

making the power supply choke input rather than cap input

making the rectifying caps bigger or smaller

replacing electrolytic caps with film (polyprop)

replacing silicon rectifiers with faster ones, and then with a tube
rectifier.

All of these affect the quality of the sound, and its presentation
quite fundamentally. It teaches a basic principle of amp design
unfortunately rarely discussed in the irrelevant welter of discussion
of date codes on Mullard tubes, that the power supply is the most
important part of the transfer function of the complete amp.

Moving on to more sound-shaping, in the signal section, you can also
try:

rewiring for a higher output impedance if the transformers allow it

making a makeshift reflected primary impedance change by mismatching
speakers to output taps

fitting better quality transformers

changing the quality of coupling caps (they should of course all be
film)

changing the value of coupling caps after a study of the Miller Effect


changing caps in the cathode circuit

All of this teaches and demonstrates circuit theory, including
subtleties about the actual load on signal tubes as distinct from the
apparent one.

A very fine tube study tool often overlooked is the computer already on
your desk, on which

with in any drawing program you can put up transfer curves (plate
current vs late voltage vs negative grid bias with the signal level
superimposed along the slope of the output transformers primary
impedance, from which you can calculate output power and the frequency
distribution of distortion

with a spreadsheet (Excel is common and good), which you either build
yourself from the formulae in the RDH (see below) or find from Jim de
Korts DIY site, or that of the excellent John Brocic (Al, whats the URL
of his site?), or that of the superb Steve Bench, study changes in all
other parameters

as an example, you can determine the minimum of negative feedback with
which your amp will be stable (see RDH below), and then by soldering
alteration determine if lowering the NFB sounds better

All of this is very useful for understanding complicated interrelations
which could be expensive or impossible to study on real amps without a
whole plethora of measuring equipment.

At this point you can move on to swapping tubes:

fit a beefier driver to deliver more current to the grid of the power
tube requires fundamental redesign, of which you should by now be
capable

change input and driver topologies first to SRPP and then to anything
that takes your fancy

try input and interstage transformers from surplus or cheap
(Southwestern???) sources

try inductive constant current loads either by wideband chokes or by
faking it with silicon (see Steve Benchs site or old copies of
Bottleheads Valve magazine)

swap tubes because some types are fundamentally more precise or sweeter
than others (6SN7, 417A)

discover that 12AX7 belong in invisible current pre-amps and 6DJ8 is
for poseurs with cloth ears

swap tubes of the same type for better brands if you want to be a
fashion victim

All of this teaches that within very wide margins the quality of an amp
depends less on the tube types or brands it is built with than the
thought with which it is designed, and the determination with which it
is developed.

Finally, you can make truly fundamental changes, which may require more
sensitive speakers because they trade off power for better quality
sound:

change a Class A/B output amp to operate in Class A1

change pentodes to operate as pseudo triodes

fit real triode output tubes (may require an airgapped output
transformer see also Bottleheads sit for shunt power feed alternative,
which he calls parafeed)

You can also develop your amp to a lesser extent, once you have
everything else as you want it, by replacing ancillary components:

a good pot for a cheap one or even a stepped attenuator

good quality connectors and switches

ceramic sockets with connectors plated silver or gold

This is mickey mouse stuff. There is even pickier stuff. I use Cardas
wire, for instance, because I can afford it and find George Cardass
golden section ideas aesthetically pleasing, but in general, if the
wire will carry the current and is rated for the voltage at 70 degrees
or over, you wont hear any difference. If the amp is wired already with
a reliable brand of UL1017 (someone straighten me out if that isnt the
600V 70 degree spec), you wont hear any advantage for rewiring with
something more expensive. All copper wire starts out oxygen free all
wire starts absorbing oxygen the minute it is cut. There are some minor
exceptions that dont affect consumer audiophiles: On very long runs of
speaker cable there is an audible distinction between using Cardass 5TC
and the normal audiophile figure of eight crap sold for more per foot
by others. The biggest difference you can make with wire is by simply
shortening your connection leads.

By the time you worked through all this, if your ears are good and you
are not impressionable (i.e. not a fashion victim or too impressed by
your own cleverness) you will understand the first thing about amps,
which is that a well amp is in balance with itself, its sources and its
speakers, and of course with you. There is no single parameter of
goodness, and there is no better judge of quality hi-fi engineering
than your own ears. It is unfortunate that this is an expensive and
time-consuming journey, but arrival is very rewarding indeed. I wish
you luck.

The book you should have above all others and study religiously (Im not
jokingI have a copy in each of the lavatories in my house I use, and
several others in at my design desk and worktable and so on) is F
Langford-Smith biblical opus, the Radio Designers Handbook, published
in facsimile of the last tube era edition by Butterworth Heinemann,
shorthanded on newsgroups as the RDH. Earlier it was called the
Radiotron Designers Handbook, but secondhand copies are as pricey as a
new one. Next to your soldering iron, the RDH is the best investment in
your tube education you will ever make.

HTH.

Andre Jute
Visit Jute on Amps at http://members.lycos.co.uk/fiultra/
"wonderfully well written and reasoned information
for the tube audio constructor"
John Broskie TubeCAD & GlassWare
"an unbelievably comprehensive web site
containing vital gems of wisdom"
Stuart Perry Hi-Fi News & Record Review