|
Summer
2009
I am a slow learner and I have the worst memory in the world, which may be an asset but it sure is embarrassment. I sound like a darn fool when discussing mathematical problems. I continue to ask the same dumb questions and rely on the advice and consul of guys like Steve Berger, J.C. Morrison, Komuro, Andy Grove, Kondo and Nobu Shishido. Without their help this amplifier would not exist. On close examination, the only real talents I can muster are a basic ability to paint pictures and a potential to (sometimes) be quiet and listen. My (imagined?) capacity to glimpse a painting, read poem or visualize the forces of nature is the source of my belief that I can indeed design audio amplifiers.
The Secret Is Forgetting In art, the simplest ink sketch — even if seemingly very crudely executed — can stimulate deep feelings and engage the focus of our whole being. Amplifiers must do the same and present music as if we were facing our own personal nature directly. When we listen to music in the home we should feel every pulsation of it as if it belonged to us. Reproduced music attracts our attention most easily when we can "identify" our own nature in it. Therefore we should let our own character resonate in the design of the amplifier. When the new amplifier is very good, the music comes from the speakers like ‘the breath of life". Music reproduction should feel direct and be perceived as having a similar substance and material character as a live musical event. The presence of the musician should appear before us.
Working In The Lab For me, the process of designing an amplifier starts with making pictures and setting my mind at ease. I need to feel I can do this. I imagine a nearly invisible, vibrating, electromagnetic slipstream forced to pass through a series of lenses and I suppose that I can compare the image of the original on one side to the thinner, bigger expanded version appearing at the other side. I try to keep a dynamic energy model in my head. I think power and energy not sinewaves. I struggle to visualize the effects of a series of partially mirrored pieces of glass (filters) on the energy. I picture a river of energy impressed on the load. I try to invent lattices of components that do the least damage to the original proportions of the energy stream. The components, the tubes and the transformers must be in good alignment. They must harmonize. I look for an ecology, a synchronous structure, a simple mathematical model and the most benign arrangement of all the physical aspects. This kind of mental imaging is, I believe, critical towards making quality amplification.
Blink, Blink, Blink... The Feral Eye is a concept amp. It was conceived by me but several of my friends helped me bring it to fruition. Feral Eye is really an amplifier format designed for home builders as an inspiration towards their own creation. Look at its architectonic aspect. It represents my thoughts about what it takes to make a push-pull amp sound less shattered and artificial. It comes out of a desire to look again at an underdeveloped technology and see if something better can be done with center-tapped primaries. Before I ever built a SE amp, I had a "paradigm shift" while listening to records at a friends house. This unusual friend had an AR turntable, Decca cartridge, RCA type preamp and a little "Bud box 2A3" amp driving Western Electric 755As. When I asked him why I loved the music so much today he said, "The only natural way to split phase is with a transformer or a splitload inverteri". Looking closely at the Bud box I noticed the UTC LS-55s and a UTC interstage transformer.
Several Years Later... The Feral Eye amplifier is a topology that can be used with almost any Class-A power tube. Believe me, you have not heard how psychedelic and colorful the 6L6 or the EL-34 can sound until you try them with this configuration. Using the "cathode feedback" secondary coil hookup with the above pentodes (in triode) will give you an amp with an extraordinary high frequency integrity and the kind of bass you never thought was possible with tubes.
Summer Of '97 The following are some of the poetic ideas that informed this design: 1) Keep resistance values low. If you can replace a resistor with a piece of iron and a coil.. do it~ Resistors dramatically effect tone character and resolution. The worst offenders are high value plate load and cathode resistors. This is not simply an issue of metal film vs. carbon composition. It is the value! Using plate resistors of greater than 2xRp, despite issues of measured THD, make triode gain stages generalize and sound rough and jumpy. The trick is to reduce the value of plate resistance to the point where the transconductance (and probably THD) are maximized and then reduce it a touch further.
Choke Is Good 2) Use low mu triodes. (This should have been number one.) Low mu and plate loads between 1x and 2x Rp give whole, complete, unedited type reproduction. 3) Avoid differential amplifiers. My friend was right. Transformers and for those on a budget, split load inverters, give a push-pull amp a more "together", less shattered sound character. Transformer coupled amplifiers, with quality iron and well calculated time constants, sound considerably more intact and continuous. The most noticeable difference between live and reproduced music is that the real thing doesn't sound like it has been taken apart and put back together. Hi-Fi as most of us know it has that "Humpty-Dumpty" sound. 4) Keep the parts count minimal. The transformer as phase-splitter and driver serves several very useful ends. In this design it replaces at least four resistors and two capacitors. It allows for greater volt 5) Because of the zero crossing inherent to PP designs I probably should have called this amp The Blinking Eye. My basic feeling about push-pull vs. single-ended is that a PP amp designed by a perceptive, discriminating designer will always sound better than a SE designed by Mr. Radiotron. Bottom line, I am certain that SE has certain very tangible advantages when listening pleasure is the main criteria. PP has a slight edge when it comes to adaptability. 6) Use choke input filters. To me they give a sense of relaxation to the music making. They take the virtues of tube rectification one step further.
Courage Under Fire Connecting the filament transformer center taps to this coil not only reduces distortion at the frequency extremes and lowers the output impedance but it also allows the designer to use A/C filaments in push-pull with very low values of A/C ripple left on the output. To hook them up correctly it is important to measure the gain of the amplifier with the heater trans CTs connected to alternate sides of this coil. The connection with the lowest gain is the correct wiring and the one with the higher gain is the positive feedback version. The driver stage, with the Tango NC-14, can alternately employ the 6L6 or 6V6 tubes. My personal favorite is the 6L6. Go mad. Try the 6L6 driver stage and a 6L6 (ultra linear) output with your Quad ESLs. Do not let the current get too high on this stage or the NC-14 core will saturate on climaxes. The 6J5 is perfect with the TC-160. I love this stage! It is so good you can use it to improve almost any amplifier. You can use other brands of iron. But! If you lose the secondary feedback coil you might as well switch to single-ended. The beauty of this design is in the use of iron. I promise you that, if you build an amp, with this arrangement of metal, you will experience a new type and quality of music reproduction. The Feral Eye plays music very differently than most of us are likely to have heard before. It sounds more full, whole and complete than any traditional PP amp I have experienced. It is very rich sounding. It loses a bit to quality SE amps on pure magic and texture but makes up for the loss by driving a broad range of loudspeakers. Regard it as a modest gift from the New York City Triode Mafia. Click here for a pdf of the schematic.
Get the entire collection
of Sound Practices now!
|
|
||||
I have a confession to make.
By any rational standard, I am not qualified to design an audio
amplifier. I have built every feasible circuit. I have read dozens of
books and even studied in the classroom of my mentor Arthur Loesch.
Still, after almost twenty years of trying, I feel radically unprepared
for the chore.
