Ralph Karsten of Atma-Sphere joins us today for an exclusive interview to discuss his company’s unique OTL (output-transformerless) tube preamplifiers and amplifiers, which were recently shown at RMAF 2009.
Atma-Sphere has been in the high-end audio business for over 30 years. What were the origins of Atma-Sphere? To what do you attribute your longevity?
I founded the company to build a new kind of OTL that seemed to sound better (I had been building them for a few years before founding Atma-Sphere in 1977). Reliability, performance and taking the time to be accessible to customers is why we are still here.
From your perspective, how has high-end audio changed over the years?
The brick-and-mortar dealers are in trouble from the internet. Auditions will be a lot harder without them! Tubes have gained a lot of ground; there are more tube equipment manufacturers in the US now than in 1958.
From its inception, Atma-Sphere has offered output transformerless (OTL) tube amplifiers. For the uninitiated, what is an OTL amplifier and how does it differ from other tube amplifier designs. What are the advantages of OTL amplifiers?
We have a paper on our site that gives more depth. OTL (these days) refers to a tube amp that has no output transformer. The advantage is that you can get a lot lower distortion, greater bandwidth and greater detail, getting you a lot closer to the music. [/i]
Both your preamplifiers and amplifiers employ a patented Balanced Differential Design®. Would you tell us the origins of the design and its advantages?
Back in the 70s I was involved with a a lot of recording, and saw how balanced lines were an advantage to recording that somehow had not made its way into high end audio. I set my goals to change that, producing the world’s first high-end audio balanced-line product, the original MP-1 preamp, in 1989. I also recognized that a designer can get overly-focused on a certain idea or motif, so the trademark is a play on words to that effect- a ‘balanced’ approach to solving audio issues
In the past, OTL amplifiers earned a reputation for instability and unreliability. What is it about your designs that overcome those problems? Is your solution your patented Circlotronic® circuitry?
The prior art of OTLs were amps that had asymmetrical signal paths and topology resulting in high distortion. This required a lot of feedback to get linearity out of the designs. Negative feedback is a destabilizing design element in any amplifier. Our use of the Circlotron was the first of its type- a balanced differential driver circuit that allowed for an inherently low distortion design that did not need negative feedback to be linear. The result was an amplifier that is unconditionally stable- it will not oscillate with any input or load condition. That was the primary barrier to OTL reliability. The other is controlling the power tubes, something that I have noticed that other OTL manufacturers still struggle with. The Circlotron allowed us a lower output impedance, meaning that a wider range of speakers is available to us, IOW 8 ohms is easy for all but our smallest amp, and our bigger amps do 4 ohms quite easily too.[/i]
Atma-Sphere markets amplifiers with output of 30, 60, 140, 220 and 500 watts per channel. The tube compliment of each of the amplifiers appears to be the same, except for the number of tubes. Does each of your amplifiers use the same circuit design?
They are very similar, differing in execution, with greater circuit enhancement as you move up the line.
Do your amplifiers present any special challenges for speakers?
No. If you want it to sound right, I think the big challenge has to do more with the design paradigm that the speaker designer is working with. You can design to look good on paper, or you can design in order to obey the known rules of human hearing. The two are quite different!
Do the same advantages of OTL amplifiers extend to preamplifiers such as your MP-1 Mk. III?
Yes, but for a different reason. In an amplifier our goal is to get rid of the output transformer, in a preamp it is to get rid of the output coupling cap, which is more debilitating for a tube preamp than the transformer is for a tube amplifier! Of course, we are also looking to produce a balanced output, so our preamps deal with both of those issues.
What improvements have been wrought in the latest Mk. III incarnations of your products?
We have reduced distortion significantly, thus improving detail and making for a smoother presentation.
Recently there has been a stir in the audiophile community over the 500 hour suggested break-in time for the Audio Research Ref 5 preamplifier. Do Atma-Sphere preamplifiers and amplifiers require lengthy break-in. If not, what break-in time in required for your products to achieve their full potential? Can you tell us what electrical and physical changes occur in the components during break-in?
We definitely experience break-in, so I recommend to people that they give the equipment about 6 weeks to break in. Its hard to say exactly how long that time really is, since some people are going to play the gear for 6 hours a night, and others only for 2. That’s why we say 6 weeks. In practice it seems that we only really need about 75-100 hours to set things right. We have been able to measure differences in power supply voltages, so we think that a lot of the break-in has to do with filter caps, although we also think that coupling caps play a role. Part of the reason for thinking this is the fact that over the years as caps have improved, we have noticed not only a shorter break-in time but also a shorter ‘warm up’ period.
I was surprised to find that you offer a turntable based on the vintage Empire 208. What is it about the Empire 208 that make it a good candidate for upgrading? Do you have much demand for your turntables?
We’ve had a lot of fun with the turntable. The original Empire was well thought out. It only lacked good resonance/vibration control, which was poorly understood 45 years ago. The Empire also has a robust drive, essential in any turntable regardless of being idler, belt or direct drive. We don’t have a heavy demand, as we (so far) are limited by the availability of the original machines for parts, but if people actually heard the finished unit that could easily change.
I noticed that you also carry a line of very moderately price interconnects and speaker cables. What is unique about you cables that make them particularly well suited for use with OTL designs?
We don’t have a speaker cable since there are some very good ones out there we don’t care to compete against. We *do* offer balanced line cables simply because the whole reason for balanced line is to **eliminate** cable artifact and differences. So a good quality but inexpensive cable should do as well as a really expensive one, and in practice we have seen an $85/25 foot set of cables be indistinguishable from a $24,000/24-foot set. This does not work for single-ended. That’s why we make balanced products, so you have the option of getting of the interconnect cable merry-go-round.
Is there anything else you would like to tell us about your company or products?
Yes- we’ve seen a lot of research into human perception of sound in the last 2 years. That research has proven what we have been maintaining for the last 4-5 years or so regarding the rules of human hearing: the more you violate those rules with the playback of sound, the more the perception occurs in the intellectual centers of the brain rather than the limbic/emotional centers of the brain. The mainstream of the audio industry has for the most part ignored the research of this type, which goes back 45 years. This has resulted in a set of design, test and measurement rules (which I call the Voltage Paradigm) that you have commonly seed in magazines for nearly the same period of time. These rules ignore the more important aspects of human perception, instead focusing on a simplified model based almost solely on tonality.
I’m not denying the importance of tonality, but the industry needs to understand that the ear/brain system is more complex than that. I for one maintain that more of these ‘perceptual errors’ are caused by distortion than by frequency response errors, and the current physiological research is backing that up. At any rate the ear/brain system is looking for other more important cues that have nothing to do with frequency response, but the industry so far has ignored this fact. The result is that there is a predominance of opinion and product that stands squarely in the way of real, measurable progress. For our part we have been looking at the rules of human hearing and designing our equipment to meet those rules; if the equipment does not follow the rules of human hearing, its not likely to sound very good either, nor will it engage human emotional response. That is to me simply common sense.
Negative feedback has been shown to violate one of the more fundamental rules of human hearing (how we perceive volume of a sound). We’ve been working on how to reduce distortion and output impedance without feedback for over 25 years. The result is something that sounds a lot more like music than the prior art. [/i]
Thank you for joining us today.
