More than a hundred years ago, a series of discoveries led to the invention of the vacuum tube and the first time that electrical power could be electronically modulated. A weak signal could now produce a more powerful analog of itself and audio amplification became possible. Vacuum tubes enabled radio, television, radar, and fast computers, but their large physical properties motivated a development team at Bell Labs to come up with a smaller equivalent device. Their successful transistor device proved to be a superior substitute for the tube, replacing it almost completely, but a certain quality of the tube defied its obsolescence. Dr. Graham Hardy, chief engineer with Marian Research, values that quality and exploits it in his high-end equipment designs. Here is a question and answer dialog we had on the subject.

When Thomas Edison was working on light bulb designs he added a separate electrode to the inside of one of his vacuum bulbs to try to solve a problem with blackening of the glass. Accidentally his team discovered that the extra electrode caused a current flow through the vacuum of the bulb, which was previously thought impossible. Do you think this was the beginning of the vacuum tube or were there other similar discoveries that should be credited?

Thomas Edison was one of many who made a contribution to the evolution of this technology. Historically, developments in electrical engineering were inextricably linked to the progress being made within the physical sciences (mainly physics). If we look back at the period from the last quarter of the 19th century to the first quarter of the 20th century, during that ~50 year period, there were some crucial discoveries made. Frederick Guthrie discovered thermionic emission in 1873. Thomas Edison in 1883, while investigating the properties of light bulbs, observed the current flowing between his filament and a 2nd element and found a rectifying function (called the Edison Effect). He patented his discovery, and although intrigued, he didn’t see a use for it. A significant event in 1897 was the discovery of the electron by J.J.Thomson. Somewhat later, Ambrose Fleming made a true rectifying diode, based on thermionic emission, while trying to assist Guglielmo Marconi with his wireless telegraphy experiments (1904). Lee de Forest added a 3rd electrode in the form of a fine mesh grid in 1907 (thus giving us the triode). The first amplifier using a vacuum tube appeared around 1911. Irving Langmuir developed true ‘high-vacuum’ tubes in 1915 which allowed much more efficient cathodes to be introduced. In order to suppress high frequency oscillations in vacuum tubes a 4th electrode was added, called the screen grid in 1926 (thus giving us the tetrode). The addition of a 5th electrode came about in 1929 (thus giving us the pentode). Also, at this time the indirectly heated cathode was introduced, thereby permitting AC current to serve as the heating source (instead of a battery). Finishing off this period, by the late 1930s there were literally many hundreds of different tube designs in use.

Can you briefly explain thermionic emission?

Thermionic emission is the phenomenon of electrons being emitted from the surface of a metal as a result of the transfer of thermal energy within the metal lattice structure to a small number of electrons (note: this classical way of thinking about it suffices for our purposes, however a proper mathematical treatment needs quantum mechanics). The degree to which a given metal surface can produce copious emission varies greatly between different metals. The surfaces that have been found to work well are: tungsten, thoriated-tungsten and oxide coated cathodes.

At what point in your career did you decide that you wanted to work with vacuum tubes?

Well, in the early 1990s I read a number of reviews in high-end audio magazines of amplifiers being designed using vacuum tubes. In a sense, this period represented a Renaissance of sorts for a technology that had been supplanted, decades before, by the development of the transistor. I was fascinated by the way reviewers marveled at the sound made possible by these devices, and intrigued as they struggled mightily to analyze and communicate to the reader the exact nature of the improvements wrought by the rebirth of this old technology. I was also amused by the fact that these amplifiers, when measured on the test bench, seemed quite inadequate when compared with their modern solid state counterparts. In essence, the solid state designers were going to great lengths to provide great sound and great bench performance in their designs, and all the vacuum tube designers had to do was build amplifiers using quite mediocre circuits inherited from the 1950s and 1960s. It bothered me that the power amplifiers were still being designed on the assumption that the end user was willing to spend time biasing each output tube using a meter and a screwdriver, while risking getting burned, and would have to do this repeatedly throughout the lifetime of the product. Around 1994, I knew I wanted to acquire powerful vacuum tube monoblock amplifiers, with the bench performance of a modern solid state amplifier, and requiring no user adjustments (other than replacement of the tubes). I also wanted the vacuum tubes to be completely exposed so one could enjoy the (considerable) visual beauty of these wonderful devices. Moreover, the cosmetic appearance of the designs at that time was unattractive, with their cheap sheet metal construction and "pro-audio" styling. Frankly, when I looked around, no one was building what I wanted. So, I was motivated to design and build them myself. To cut a long story short: this was the birth of Marian Research.

What qualities do vacuum tubes possess in comparison to solid-state devices that make them desirable for audio equipment?

The transfer function of a vacuum tube is very linear, so it is possible to design a power amplifier having low distortion without using large amounts of global feedback, which we now know is responsible for a lot of the ‘hardness’ present in the sound of high feedback solid state designs. In addition, vacuum tubes are high frequency devices. The capacitance of the control grid for a large beam power tube is about 15pF. The capacitance of the gate on a power FET might be 250pF or more*. These vacuum tubes are the true high frequency champs of the amplifier world! The combination of linearity with great bandwidth makes for intrinsically natural devices for audio reproduction.

* This effect comes about because the output impedance of the previous stage forms a low pass filter with the grid (or gate) capacitance in question, and thus limits the high frequency bandwidth.

Can this difference be quantified with electronic equipment, such as an oscilloscope?

It depends. Some subjective differences between the sound of tubes vs. solid state electronics can be measured, but many cannot be, at this time. That doesn’t necessarily mean that we won’t be able to measure them in the future. The performance of different devices depends to a large extent on how they are used in circuits. Thus, it’s possible to get subjectively "bad sound" from either type of device if the design is poor. We know that different listeners have different preferences or priorities in evaluating sound, which also relate to the types of music and instrumentation they enjoy. While test instruments and methodologies have become much more precise, our level of sophistication in terms of subjectively identifying both faults and superlatives has increased with the availability of vastly superior digital source recordings and remastered analog recordings.

Years ago I read a magazine article about tube amplifiers. The article suggested that the better sound was from the way the tubes "colored" or "mellowed" the sound, almost like whiskey aging in an oak barrel. You're saying that it's actually due to the better accuracy of reproducing the audio signal?

Any comparison requires some sort of reference. Years ago, that reference may have been the early solid-state amplifiers that sounded harsh. Today, we have superb sounding amplifiers using tubes, solid-state devices, and hybrids of the two. No matter what devices are used, the performance of high quality modern amplifiers is a quantum leap beyond that of older designs. If you perform blind A-B tests of an old tube amp versus our modern tube design, the difference is striking — ours is far and away better sounding.

Old tube amps may sound "mellow" because they are smearing or blurring the detail ("rounding off" transient waveforms), but the tubes are not necessarily responsible for that. Through extended listening it will become apparent that this is not actually better sound. In fact, it is very unsatisfactory in reproducing contemporary music styles that utilize electronic instrumentation, such as electric guitars, synthesizers, and synthesized percussion. Modern recordings of all styles have much greater dynamics than older ones. Along with the blurred detail, old tube amps add a lot of audible noise that intrudes on one’s enjoyment of the music. This can include "snap-crackle-pop" during warm-up and a copious amount of "hiss" that impinges on the dynamics and destroys the "black velvet" background that all of the sound is built upon. Older tube amps also have fairly high harmonic distortion and transient overhang or "ringing", which may sound pleasant at first, but soon it becomes tiring and it’s obvious that a veil has been placed over the music, when compared with the neutrality of our modern design. It’s a bit like wearing rose-colored glasses that distort the world by concealing its full spectrum.

Our sophisticated tube amplifiers have maximized the qualities that everyone enjoys: a liquid, spacious, timbre-accurate, and dynamic reproduction of the music, a silent background, and no fiddling with bias adjustment. Who wants "mellow" when they can have "exquisite"?

I still remember the tube testing machines that they used to have at drug stores. When our TV was acting poorly my father and I would pull out the tubes, go to the drug store with them, and test them out on the machine. There was a large dial on the machine that indicated "good" or "replace". What care and maintenance is required for a modern vacuum tube amplifier?

The enemy for vacuum tubes is excessive heat. For long service life a vacuum tube should not be operated outside the tube manufacturers’ data sheet. The most important parameters are filament current and plate dissipation. Also, high power output tubes should be operated with adequate spacing between each tube so that each tube is not radiating too much heat into its neighbors. Of course, most of these things are set ‘in stone’ by the designer and are not under the control of the end user. At Marian Research we use quite conservative operational points for all our vacuum tubes, so as to promote reliability and long life.

On the business of ‘caring’ for your vacuum tubes, there is one thing you can do to improve the sound quality of your amplifier(s). All vacuum tubes exhibit, to some small degree, the phenomenon of ‘microphony’ (note: there is even evidence to suggest this occurs to some extent with solid state devices). Microphony exists because the delicate electrodes in the tube vibrate due to resonance effects as they ‘listen’ to your loudspeakers. Fortunately, one can purchase after-market tube dampers which attach to the glass tube envelope and thus ameliorate these subtle colorations. The addition of these devices causes the soundstage of your system to gain a quite noticeable improvement in ‘focus’ and in the realism of ‘transient attack’.

What is the availability for replacement tubes if needed?

Vacuum tubes fall into two categories. There are ‘New Old Stock’ (NOS) tubes, which are new in the sense that they have never been used, but old in the fact that they were manufactured decades ago. Then there are vacuum tubes currently being made at factories outside the USA, principally in China, Russia and Slovakia. At Marian Research we obviously cannot rely on a finite resource like NOS tubes, so we use current manufacture tubes from around the world. Summing up, I would say the current availability of vacuum tubes is very good and can be expected to continue for many years to come.

In what ways are you able to further enhance the qualities of vacuum tubes using modern technology?

I would like to refer you to a good discussion of this at the ‘Design Philosophy’ page at http://www.marianresearch.com/philosophy/

What types of loudspeakers are compatible with your amplifiers, and what should the considerations be?

Most loudspeakers having an impedance of between 4 & 8 Ohms will prove compatible with our amplifiers. However, if one is striving for superlative sound then the choice of loudspeaker becomes more complicated. Although we can state that our amplifiers are wideband, an even more important consideration for me, as a design engineer, is time domain performance. By that I mean, how well does the amplifier reproduce a short pulse. Our amplifiers are optimized for this. By extension, making use of good amplifier pulse performance requires that the loudspeaker do essentially the same thing. Unfortunately, most loudspeaker manufacturers do a poor job of documenting time domain performance in their published specifications. However, the sales folks at Marian Research can provide some good suggestions here.

Does one impedance type perform any better over the other?

Not in a way that I could characterize as a preference based on sound quality. Our power amplifier will deliver considerably more power to a 4 Ohm speaker than one having an impedance of 8 Ohms (~60% more).

Do you have a preference for ‘acoustic suspension’ or ‘bass reflex’ type loudspeakers?

No. I have heard very good implementations of both types of design. My personal preference is for large planar dipole loudspeakers. At Marian Research we use acoustic suspension, bass reflex and planar loudspeakers to baseline our products.

Do you have any future plans for yourself and your company that you’d like to share?

We are currently working on an ‘AC regenerator’ designed to take the unregulated ‘dirty’ power from the wall socket, and transform it to provide perfectly sinusoidal, tightly-regulated power to your stereo system (debuting in the fall of 2009). Stay tuned!

electricalfun.com, ElectronCafé.com, and WorkbenchFun.com are trademark properties of ElectricalFun Media
Copyright © 2009