A U
D I O... P A P E R
A SINGLE
ENDED TUBE AMP FOR THE MASSES
by Steve
Deckert
August 1996
A
design log that was written during the development
of the Original Zen Triode Amplifier
August
1996
High Fidelity Engineering (Decware) has a clear focus that revolves around preserving
the forgotten past of hi-fi in the 1960s before it took a major
step backwards from the perspective of sound quality. By that I mean
that since the invention of non-discrete IC based solid-state
receivers, overall sound quality has declined when compared to the
older tube counterparts. Believing this gets harder and harder with each generation.
Kids don't know what records are, and they think tubes are light bulbs.
And quite frankly, it is completely impossible to relate to the
kind of sound quality I'm talking about unless you've been exposed to
it. Kind of like a man trying to imagine what it must be like to be
pregnant and give birth, something he'll never know... at least not in this lifetime.
If
you're the average person with a stereo - even a fancy new
surround-sound stereo then you enjoy the company of approximately 98% of
all the people with stereos... and I'm sorry to say that
presenting this amplifier in a way that you can grasp (because it
goes against everything you know) may be no less complicated than a
pregnant man.
The
passion behind this idea for me comes from the desire to expose the
common man owning the common stereo - to the magic of triodes, with a
secondary objective of messing with the common audiophile who believes
specifications are the 11th commandment. Someone should have done this
long ago, but because of the unorthodox thinking required to pull off
such a feat, I don't suppose other companies wanted to take the risk. For example, if I tried to market a 6 watt amplifier with around
10% distortion who would take it seriously?
An entry level "
audiophile " amplifier that would be affordable to the masses has been
my dream for many years now. To find a way to make it affordable and
make it appeal to the average person is obviously the difficult part.
The passion to do so comes from past personal frustrations of not being
able to afford what I want and not being able to enjoy what I have
because I've heard what I want! (been there haven't ya!) I can assure you most of us have a hard time justifying the crazy amount of money required to obtain a true high-fidelity system.
The
general perception of an audio system to the masses is pretty much
anything you could find at a " digital ready " electronics supermarket.
Well guess what... the masses all have ears as good as any audiophile
and the ability to appreciate the same things that result from high end
listening. Those who've never been exposed to three-dimensional
high fidelity can hardly be chastised for not appreciating the
difference between the two, or for not being aware that there is a
profound difference.
The science
of sound and harmonic resonance is far and away the most
complex and least understood science on earth, in my opinion. I
would say that the monumental joint effort of all the people, and
the equipment required to send up an Apollo mission would be
simpler than building a perfect playback system. It takes
significant effort in design and a significant quality in parts to
achieve good sound. It should be pointed out that the HIGH-END
audio market is oppressed with the monumental and
expensive task of improving the high-fidelity playback system
we commonly refer to as the " stereo " and that is why a
respectable stereo system costs between 15 and $30,000.00 Gee,
that's about the price of an automobile!
September
1996
My objective
with this design is to get the absolute highest
quality… harmonically in-tact sound in the hands of the wanna be
audiophiles. My research and experience over the years has taught
me that tube circuits are the way to obtain this objective. If I
were designing this on paper with specifications in mind I would
have chosen solid-state circuits. I really prefer the natural
sound of tubes and while I have heard some solid-state solid-state amplifiers that I could enjoy, they start at about $7500.00.
The market is SATURATED with solid-state gear, and as I said, I don't feel it is the format necessary to complete this goal.
Up
until this summer, I have been building Class A1 push-pull tube
amplifiers of various designs with the intention of marketing one as
the entry level amplifier. The part I've been wrestling with
(like all small manufacturers) is keeping the cost down and
yielding terrific sound quality. Success usually seems to go to
those who have achieved the correct balance of compromise
(understanding of course that everything in audio is a balance of
compromise).
The
push-pull tube amp designs seem to be the most popular because they
can be fairly inexpensive to build, have plenty of power, and sound
quite good when compared to mid-fi solid state gear. I really thought
the answer could be found in a push-pull design for those reasons.
This
summer while playing with single ended tube amplifier designs for my
own personal stereo, I stumbled into some results that forced me to
take a real second look at exactly what are watts? You know you read
these ads for those five watt triode amplifiers that start at around
$5,000.00 and go way up from there and wonder why would anyone pay that
much for such a thing if it only has a few watts?
Lets just
say some enlightenment from the audio spirits came upon me, and it was
enough to realize that at a normal listening level, most of
the musical content can be found in the first magical watt of
power. I then realized that the priority in high-fidelity
reproduction should be focused on that first watt. I have been trying to
truly understand this for some time now, and in particular, been
trying to define in my own research the reason why solid-state
watts seem less than tube watts to the ear, yet equipment measurements would indicate that they are capable of achieving the same amplitudes.
Solid-state stereo
gear has a tendency to sound thin, and quickly run out of headroom
(clip) when pushed. Tube amps are very different. If you compare a
40 watt tube amp with a 100 watt solid state amp or receiver, the
tube amp will put more music in the room, and get louder every time. You
will find that at nominal listening levels, the loudness button
is needed to get the solid-state amp to sound full-bodied,
yet the tube amplifier sounded warm and full with a dead flat
signal. This is a great example of how " watts " are not " watts "
and a prelude to a secret only the most advanced audio gurus will
share, and that is that specs in audio gear mean nothing.
Why
is it that a 10 watt musical instrument amplifier such as a guitar amp
will in real life (and on stage) get loud enough to split your brain
in half, yet it seems to take mega bucks and major stereo gear with
100's of watts to reproduce the same sound in your living room, a room
that remains a fraction of the size of a live performance? Perhaps it's
because everyone is going about it wrong, confusing convenience with
performance.
October
1996
Enough
rattling on... Two weeks ago I completed prototyping the circuit for
a single ended low power tube amplifier and have been listening to it
ever since. The schematic is at the top of this page. At this stage
I have already decided that this will be the chosen design for the project.
The actual cost will decided by the cost of the output transformers.
One of the main reasons single ended tube amps are so much higher in
cost than their push-pull counter parts is that the output transformers
are completely different. In a single ended design, the output transformer
must be designed to handle the DC current at the bias point, so a special
transformer must be used. It features an air gap that optimizes the
coupling at low frequencies and the DC current that serves to lower
the permeability of the core. Without the air gap, the iron will saturate
under too little DC bias to accommodate the needs for a single ended
triode. Too much gap will reduce primary inductance so that the lowest
frequencies will not pass without attenuation.
All
other things being equal, the output transformers (or IRON as I call
it) has the final say in the resulting sound quality. When you're looking
at different iron for a design, you find that the standard push-pull
output transformers range in price between $75. and $400.00 ea. (per
channel.) and you find the single ended output transformers range in
price from $150. to 1200.00 ea. That BTW is why the single-ended stuff
is so expensive.
I have
a 50 watt (ea. channel) tube amplifier that I built up to a
reasonably impractical extreme, and I have been using that as my
personal reference piece. It powers an efficient pair of speakers
in my main listening room (over 90dB) and in all honesty (with
feedback off) sounds better than any other push pull amplifier I
have compared it to. In fact, a version of it was my original
idea for this project, and has been for years because I liked the
sound so well…
My specific design goal has wavered a bit in the past months as I
get ready to do this. If I am going to market a tube amp to people
who have never had the joy of listening to one, and given the solid
state, cranked up with the loudness on and tone controls engaged
listening habits of those people, how should it sound? Do I go for
power so it will stomp their past systems, or do I go for
pleasure so it will reveal to them the inner levels of music?
In other words, do I give them what they already have but just a
lot better, or do I give them an opportunity to discover a magic in
music that they are unaware of. It really gets into a psychology
issue, one that I have pondered for almost 10 years now. From a business standpoint, I would make more money with the prior.
This is what happened... to completely solidify the
decision. My new little single ended amplifier will run in either
Pentode or Triode mode. In Pentode it benches 5.7 watts, and in Triode,
it does 1.8 watts RMS pure class A per channel. Because of the
front end, and additional gain stage in the design it is possible to get
louder than you could ever believe is possible with 5.7 watts.
Anyway, ever since I switched the little guy over to triode mode, I
have not had any desire to switch it back. It is the most natural
real sound I have ever heard in this house. And the eerie thing
is that it achieves a nominal listening level high enough to be
exactly the same listening level I have been accustomed to.
That’s 1.8 watts Vs. 50 watts which is the same as a hundred watts
or more in a solid-state receiver. The enormous improvement
in quality has made this new little amp my full time personal
listening amp. My good ol’ favorite just got bumped. As for the
psychology issue, I will be going for the magic.In
my observations over the years the reason people turn the
volume up
to the levels they do, is to gain the effect of physically
feeling the
music. This effect is the motivating reason for turning it on,
and this
effect is the ONLY effect that the quality of equipment has to
offer
so it's no wonder. Once you have spent an evening with premium
gear
like this little triode amp, you find that the physical effect
you were
accustomed to happens sooner and at lower volumes because the
even order
harmonics are in-tact and free of odd order harmonics found in
solid-state circuits. Then the big one hits you, another more profound
EFFECT
happens in addition to the physical effect -- emotional
effect. So you
have one that strokes your body, and one that caresses your
inner self
adding a new found joy to the experience of listening to your
stereo.
November
1996
Having
my strongest talent in speaker design I have thought long and hard about
the statement: "Your speakers have the most effect on how your stereo
will sound - replacing your speakers first gets the most improvement."
Being a speaker designer the temptation to accept that has always been
strong, however I have been slowly and consistently disproving that
to myself year after year.
It
is in my opinion the amplifier that makes the most difference in the
sound of your stereo, and then your other electronics. I can safely
say that your speakers are probably the least contributors to the sound
rather than the most if your listening to good tube gear. It is not
however usually NOT true when listening to mid-fi solid state gear.
If you gave me a choice of a good tube amp and a pair of Bose speakers
or a mid fi solid state amp and a $24000.00 pair of Wilson Watts, I
would choose the Bose and the tube amp because it would sound better.
I know blanket statements like this are hard to digest, but I believe
it to be true from my own experience.
There
are things in this design that some engineers would whine about, perhaps
lots of things, but I did what I did... to achieve the desired sound.
Audio is so much more complicated than a square wave response on a scope.
I feel a significant part of the complexity can be found in the topology
of the circuit itself. Understanding that all matter in the universe
resonates, including conductors such as wire, adds an exiting new level
of depth to it. I should imagine a circuit would not be unlike a piano
with all of its notes. Combining the right ones (based on their fundamental
tones (resonance) in a way where the harmonics of those resonance's
are complimentary determines how pleasing the outcome is. Since on the
bench there is no real way to measure molecular resonance or see its
effect on the flow of electrons it is impossible to know exactly what
is going on. If we could become electrons and jump in for the ride,
how smooth would it be? Would it be violent? How many times would you
hit you head? And when we both popped out the other end of the circuit
at the approximate same time would one of us be battered up and the
other one no worse for the wear? And if so why? These are the things
that intriguing me. Now that you have an idea where I'm coming from,
I will spend the remainder of this log going over the design itself
with the intent of helping anyone who cares understand the reasons why
I did the various things I did with this design.
A
SINGLE ENDED TUBE AMP FOR THE MASSES
Part
II - The Design
November
1996
If
your not an amplifier builder this may become laborious reading after
awhile, but initially you should understand that the audiophile's fascination
with sound stage and imaging (the motivating factor for all tweaks)
is well supported in this design. I have found that the determining
factors that map your systems sound stage and control imaging are mostly
if not completely governed by impedance matching. Finding the exact
right setting on your pre-amp volume control for best sound could be
an example. When you break-up a pre-amp and amplifier into their respective
gain stages and begin to play around with the input and output impedance
of each you start to see why no two combinations of amplifier/pre amp
will ever sound the same. This is the reason for the large concern about
finding "good matches" in equipment and why some good gear sounds bad.
To all "miserable" audiophiles that have found out the hard way that
money alone can't solve the problem, I would say in fact that this is
the problem.
This
ability to tweak is what I feel it will take to win the hearts of the
serious audiophile and what it will take to be compatible with the expectations
of the solid state conditioned masses. This is not because it has millions
of bells and whistles (and it doesn't) but because the operator will
be able to "discover" the best sound by changing the configurations
of the amplifier. It makes sure that whoever buys it is sheltered from
the alternative which is taking your chances with odds of getting just
the right impedance match between all of your components.
The
schematic above is what I have prototyped and am currently listening to.
The final version will be very similar to this. As you can see the amplifier
has 3 gain stages, or optionally two, depending on where you select to
have the input go to. In the two stage (direct) mode, the single passes
through one resistor and one capacitor the tubes and the output transformer.
I'm sure you'll find merit in this "less is more" approach when you listen
to it. The additional stage making this optionally an integrated amplifier
for all practical purposes gives additional impedance balancing and more
gain so that you can listen to it at louder volumes.
The
tubes I am using are a 12AU7 and 12AX7 for the first two stages into
a pair of 6BQ5's (EL84's) for the finals. I chose the single EL84 because
it is a scaled down version of the popular EL34 but in my opinion has
a better sound. The single EL84 uses less current which is making this
amplifier a cost effective reality to build. I think it has better sound
than larger tubes because the plates are small and tightly packaged
around the screen grid meaning there is less distance for the flow of
electrons between the two. The other tubes are perfect for this design
because they are the most popular of their kind and the easiest to find.
This means the guy that buys it can play musical tubes by trying several
different brands without spending any real money.
The
biggest feature will be the switch for changing the amplifier operation
from Pentode to Triode. If you need more output, and like you music
to sound a little analytical with better specs you can run it in pentode.
If you want to here a slightly warmer more pure musical sound with less
power switch it to Triode. Obviously there are several combinations
you can get from combining the features of this amplifier each with
it's own sonic signature. You surely will be able to find one that you
love.
The
outputs on this amplifier will run most any impedance of loudspeaker,
6 to 8 ohms being ideal (at least with the current iron I plan to use).
The outputs can bridged by strapping the positives of both channels
together if you happen to have two of these and wanted to run them as
mono amplifiers, one for each channel.
A
SINGLE ENDED TUBE AMP FOR THE MASSES
Part
III - The Chassis
|
"If you're going to market this thing so that it
would appeal to the masses, it would be mistake to think the potential
buyers are going to part the living room with it like the Red Sea... Nope,
that only works when you live alone! |
Well
I have pondered the chassis design for some time now. Not only from the
perspective of cost and appearance, but from the resulting topology of
all the parts. Things like heat and shielding become large considerations
not unlike where the thing is intended to be placed in the room. Do you
make it like a conventional tube amp and plop it down on the rug, or try
to cram it into a more standard chassis like a 19" rack mount unit.
A
unique opportunity with this little amp lies in the simple fact that
it is so small. The 6BQ5 output tubes are not much larger than a normal
pre-amp tube, and have nowhere near the heat of a larger output tube.
Because of this and the size of the transformers I could build this
into a nice rack mount chassis the size of an average pre-amp (3.5"
high).
Okay,
how to give the owner of this amplifier the satisfaction of hypnotically
staring into the golden glow of the tubes... you know that's a must
- too many people talk about it. I happen to be one of them. In fact
I'm not sure sometimes if it's a vice or a gift, but I have a real hang-up
with how things look. What followed was the all too familiar, slightly
out of phase, molting period of several days when I walk around in circles
a lot. Two days ago while I was meditating to music, (I was really enjoying
some "FRESH AIRE" on American Gramophone being played on this
amp) it hit me like a Zen Lightning Bolt in the forehead! An image of
exactly what it should look like, how to lay out the entire parts topology,
and flashes of it sitting in many different listening rooms blew into
my head. I love it when that happens!
From
that image I did this conceptual drawing of it exactly as it appeared.
The picture can be seen above. The ventilation
will be through the top and bottom center of the chassis. The tube topology
will be horizontal opposing channels with the rectifier centered between
them in the back of a shielded "room" inside the chassis. I'm not sure
how cost effective this chassis would be, but I will defiantly be building
some.
A
SINGLE ENDED TUBE AMP FOR THE MASSES
Part
IV - Personal Log
Feb.
1997
I
hoped it would be ready by now, but you know how that is! Currently
I am waiting for the output & power transformers I requested that
all transformers be hand wound by one particular man, (one of the few
older fellows still alive doing it) and am still waiting. I believe
it will be worth the wait.
Last
night (this is a classic example) a gentleman stopped by with his ADCOM
amplifier to hear what it would sound like matched with a tube pre-amp.
He had the most expensive and largest amplifier they make. It looked
very nice. We listened to it for a half hour or so and he was impressed
with the idea of a tube pre-amp. Between you and me, it sounded fair
with the a few shortcomings of being a little dry and a little thin
sounding unless it was really cranked. I don't remember how much power
it is, 500 watts per side or something.
As
we finished, he asked what that pre-amp on the floor was like. He was
pointing to the SE-6BQ5a. I told him that was an amplifier, currently
running pure class A triode with a 1.8 watt output per channel. I went
on to explain the whole thing to him. Well of course he wanted to hear
it so I hooked it up where his Adcom had just been. The bass was full
& rich, and everything sounded better.
It
was having no problems maintaining the 90 dB playback level we had been
listening to and in his own words, besides saying he couldn't believe
it 42 times, he said "At this playback level my Adcom's got NOTHING
on this!" He was right.
I
am very excited about this project! With the perfectly flat hand wound
output transformers, silver wire, and 100% polypropylene power supply
there are very few amplifiers that will sound as good. The only ones
that would stand a chance are other single ended triode amps costing
as much as a new car.
March
1997
While
waiting for my transformers, I have had some time to re-analyze my thinking
on all aspects of this design. The only thing that is certain is the
circuit itself. Parts and layout are still open to review. Certainly
for the purists and seasoned audiophiles I am on track with my 100%
poly caps in the power supply and signal path. Make no compromise on
parts quality, including wire and solder and lets see how good- good
can get.
Perhaps
if I'm going to make references to "Zen" in the design of this amp I
should pay some attention to it. An interesting thing has happened this
past month while waiting for the transformers. I was commissioned by
a studio in New York to build a tube equalizer that would warm up the
recordings. It would be the last thing in the signal path just before
the master deck. They wanted that enchanting "tube sound" in their final
product, the recordings.
I
took on the project with the approach that I would give them some basic
tone shaping in the bass and treble area nothing else. I figured if
they need to EQ it more than that there is something wrong elsewhere
in the signal path. This being the case, the front end of the SE6BQ5
circuit would work perfectly. So would the basic chassis. This would
give me an opportunity to explore some variations of the project, and
let me really analyze the sound of the front end as a separate pre amp.
This also gives me an opportunity to try some different twists on the
power supply.
Thinking
about marketing and other non Zen thoughts I did a really stupid thing.
I built the full blown power supply that goes in the Zen amp for this
pre-amp project. I also made some changes to the solid state rectification
using diodes and very large filter and B+ caps. Here is a picture
of that power supply. A little overkill ha?
My
thought was build a solid state rectified duplicate of the tube rectified
power supply I would be using in the Zen amp and see which I liked better.
The problem was that I had a B+ of 435 volts and only needed around
270v. Not wanting to introduce solid state regulation (the natural solution)
I tried to wrestle it down and work with it. The result was a poor signal
to noise ratio that was audible.
I
solved this problem by coming back down to earth and just building the
proper supply to do the job. I chose two 300 ma transformers, one to
run each channel. I ran a full wave bridge off each one through a one
ohm resistor, followed by three sections of filter caps. This gave me
what I needed to accomplish a true dual mono unit. Why not right? I
did the same thing with the heaters, using one large 100,000uf 20v cap
on each side. Now I had a dual mono supply with dual mono DC filament
supply and no solid state regulation. This is an obvious improvement
over the first circuit.
The
next interesting development was the way I immediately dumped the very
best poly caps I had (combinations of Multicap and Rel caps) for everything
in the signal path. I let the unit burn in for a week and was surprised
to hear it sounding as good and actually a little better in some ways
than my pre amp. This is my own interpretation
of the "Last Pass" for the popular Dynaco tube pre amp.
The
result was that the pre amp was "too good". In other words it was very
very very fast, extremely revealing. On my high end system it sounds
wonderful. Substitute a normal CD player, or mid-fi solid state power
amp into the equation and the sound (for me anyway) was unlistenable.
One of the secrets I know to be true in all musical sounding systems
is balance. All pieces of electronic gear in the system must be of equal
speed and quality regardless of the value.
Since
I wanted to accomplish a "tube sound" in a solid state signal path of
a recording studio, this wasn't going to work. I needed something that
warms, soothes, which translates into subtle filtering. Something that
wasn't so fast and so transparent. To do this I removed all the high
end caps and installed normal mylar caps very similar in sound to original
caps used in the 60's. I also changed the resistors back to 5% metal
films. Part of me thought I was really crazy at this point.
With
great anxiety I took the reworked unit back into the listening room
and plugged it in. Get this... I was actually shocked when music came
out, and then had to quickly remind myself that of course music came
out. Just shows how brainwashed we get by listening to everybody else.
I expected the sound to be so pathetic that it would justify my original
implementation of the good caps. And it was... for the first excruciating
30 minutes... while the caps seated. But then guess what?
The
stupid thing sounded warmer, richer, and the highs were nice and sparkly
and most important, the sound stage was still there! Hmmm. Okay, we
can fix that, bring in the Sony power amp... and that 200.00 CD player
over there. Now we have a mid fi system on high end speakers, a cripple
from the start. Well, that was the best I had ever heard that gear sound.
Gee, it's working, this thing is making flat dry nasty stereo gear jump
back into the 3D high end sound stage. The difference is rather amazing.
This
means that the secret to a musical Zen amp for the masses is not going
to be found in the high end caps. (A rather hard pill for me to swallow
btw.) That means I am going to have to build it exactly like I just
did this pre amp and that means I will have to make some "Silver Edition"
units with the originally intended parts for those audiophiles who have
the gear to back it. Consider the conceptual drawing I did of the chassis
to be what the Silver or Special Edition model will look like if and
or when I decide to sell some.
Meanwhile,
staying focused on the original intent of this design I have elected
to go with a more straight forward look. And because of the success
I just had with the studio's pre amp, I am going to offer two products.
The Zen amp as intended, and a pre-amp. This way those of you who have
the larger amplifiers/less efficient speaker thing happening can benefit
from the corrective qualities of this pre amp. In other words we can
help more people create a musical sounding system. See how things work
out?
Here
is a picture of that hand wired pre amp.
August 1997
The
iron samples have finally come in and I am very pleased with them.
Now all that remains is to build the little guy and see how it compares
to the prototype.
I've
made all the decisions needed to build the first pre-production unit
and have completed it. I kept the layout symmetrical so the amp
looks good.
June
2001
Well,
what a rush! This is the first time I've gone back to this page since
August of 97. The Zen was optimized for Triode use only after
the first 25 units were sold. Shortly there after the tubes were upgraded
to SV83's and 6N1P's and it has remained basically unchanged since then.
We have hand built over 600 of these amplifiers. In fact we have developed
an entire line of Zen Triode Products along
the way. Many Thanks to everyone who is supporting us by enjoying these
magic little amps!
Steve
Deckert
NOTE:
In January of 2008 Serial #1000 of the little gray amplifier
(Model SE84C) was built and shipped. The chassis
was retired but the amplifier lives on in a slightly
larger and significantly heavier black steel chassis
with a front mounted gain control and 2 sets of inputs.
The new model is called the SE84C+. A history
detailing the revisions from the first zen to the model
pictured above can be found in the articles
section
of the web site. At the time this article was
written I never dreamed we would have hand built and
shipped 1000 of these, not to mention a few thousand
more amps that we designed based on this concept. Just
goes to show I was on the right track, and it remains
our top selling amplifier today.
HOME
BACK
Decware is a trademark of High Fidelity Engineering
Co. Copyright © 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006
2007 2008 by Steve Deckert
|