As audiophiles many of us have products we have a very particular appreciation of. It could be a beaten up old valve amp found in a parent’s attic that utterly transforms music or a long desired component that can only be obtained by mortgaging the family home. Although some take this interest a step too far and become distinctly fanatical, there are occasions nonetheless when such abundant enthusiasm is partly or fully justified. I believe the Audio Innovations First Audio power amplifier is an excellent example of the latter. Its importance is not acknowledged today, which in turn prompted me to write this advocacy about its significance historically and its continued worth as an amplifier in its own right.
Although the First Audio, and its monoblock relation the Second Audio, seem to be largely forgotten they represent a genuine milestone in audio technology. They are the culmination of a design aesthetic that emerged in the 1970’s where there was a gradual realisation that feedback can be problematic and valve technology began to be reappraised. Not only has approach to amplification radically altered how amps are designed today but the First and Second Audio have also been influential.
Introduced in 1986, the First Audio was extremely radical for that era: a high-end amplifier producing a grand total of 7.5 watts per channel! Other than a few relatively exotic Class A transistor and lower cost valve amps, virtually all amplifiers producing 20 to 30 watts occupied the budget market. Any outputting less than two figures seemed to cross some sort of psychological barrier. Somehow it wasn’t right. The fact that it was a relatively costly amp made the notion more bizarre, perhaps even shocking!
The First Audio uses output triodes. This was almost as novel as the very low output. While small signal triodes were and still are the most common valve type found in amplification, amps using output triodes hadn’t been widely available for some thirty years! Tetrodes and pentodes became the norm as they produced greater power, and in the 80’s manufacturers sought higher transistor-like output ratings.
The rebirth of triode amplification can be understood by looking to the past. Output triodes with direct-heated cathodes were the mainstay of high quality amplification in the 1920s and 30s. Output pentodes and beam tetrodes became popular in the 30’s. They are more power efficient devices as they generate higher wattage in a given circuit. Unfortunately, this comes with a much higher output impedance and significantly higher levels of distortion, including a disproportionate increase in levels of odd order distortion which is more audible to the human ear than the even order distortion that triodes primarily generate. They were used in applications where power and cost took precedence over sound quality.
After World War II negative feedback was used in amplification to reduce distortion levels. By the turn of the 1950’s with the rise of the hi-fi industry and its associated commercial pressures, output tetrodes and pentodes were increasingly used for better output power, combined with more negative feedback to reduce output impedance and distortion. Increasingly Class AB circuits replaced Class A to further increase output power but as with the switch from triodes it yielded greater levels of distortion.
Clearly high output power is not a virtue of triodes. The key reason for their use rather than tetrodes, pentodes and transistors is sound quality. This is mainly a product of their lower and more benign largely even order distortion characteristics, allowing little or no global or local feedback to be used.
Tests in the 1930’s indicated that second order distortion is imperceptible at levels up to 5% although some perceive it at lower levels. Shorter in 1950 and Wigan in 1961 published articles discussing the greater aural impact of odd order distortion. In 1957 Crowhurst referred to the increasing complexity of distortion when feedback is applied. Lower order distortion is reduced while higher order forms increase. These orders are lower in level but potentially more audible especially when modulating with the signal. These observations were not necessarily a rejection of feedback but raised issues about use. Such concerns were not widely accepted so the affair with feedback continued without complications.
By the late 1960’s, while the world was embracing the transistor and high feedback, a small movement developed in Japan comprising of enthusiasts who advocated the use of low power triode amps coupled with very efficient speakers. Jean Hiraga was influenced by the Japanese enthusiasts. In 1970’s articles of his were published which contended that the harmonics of distortion rather than total distortion levels should be evaluated. In 1986 he received an award from a well-established magazine for his contribution to the understanding of audio reproduction. That year the “First Audio”, the first commercial triode amplifier in decades, utilising neither global nor local feedback, appeared in a largely unsuspecting audiophile market.
Audio Note built a small number of quite powerful amplifiers based on the large 211 triode from 1980. The Ongaku, issued circa 1989, was sold internationally more so than previous models with the help of Peter Qvortrup, then owner of Audio Innovations. It caused a quite stir partly due to its £30,000 price!
The First and Second Audio pre-date the Ongaku by a number of years but failed to build up anything like the same level of mystique. They were far more utilitarian and cost a mere fraction of the price. In the early 1990’s the First Audio retailed for a relatively modest £1300. Its bigger brother the Second Audio (a monoblock double valve 15 watt equivalent) retailed at £2600. Both sold surprisingly well.
One gets the feeling hardly a cent was wasted in the construction of these amplifiers. They have a fairly lightweight construction for models in their price range, utilising the standard Audio Innovations art deco cabinet design common to cheaper models like the 500. Being more substance than surface, these amplifiers have their share of internal goodies. The power-supply uses a hefty mains transformer with six secondary windings. Components changed with time but handmade paper-in-oil capacitors (rare at the time), bulk-foil non-magnetic resistors, and fully hardwired circuitry are common. The output stage uses cathode (auto) biasing. Four and eight ohm binding posts are fitted to suit different speakers; early models also feature a 16 ohm binding post. A switch on the rear alters the earthing to minimise hum.
There were three versions of the First and Second Audio. The initial version (c. 1986-88) uses ECC88 (6DJ8) and PCC88 small signal valves along with 6B4G direct heated output triodes. The 6B4G is an 8 pin 6.3 volt (cathode heater) version of the 2A3 valve (2.5 volt heater, 4 pin) which itself dates to 1932. It was a popular variant used in amplification into the 1950’s but is rarely used today. The ECC88 is a double triode popular since the classic era of hi-fi. PCC88’s are similar but use a higher heater voltage.
After a short time 6B4G NOS supplies began to run low so new versions were launched using 2A3 tubes instead (c. 88-92). The company took the unusual step of having modified 2A3’s produced by Shuguang in China. The most peculiar feature of these amplifiers is the use of 2A3’s with octal (8 pin) bases rather than standard 4 pin bases. This was required since 4 pin bases were rare at the time. These octal bi-plate valves sound reasonable but don’t compare with NOS or modern monoplate 4 pin 2A3’s. Changing the bases in the amplifiers to 4 pin or buying good 2A3’s with octal bases specially fitted is an important upgrade. The PCC88’s were jettisoned for uniform ECC88 usage in the preceding stages.
The last version of the amplifiers (c. 92-96) were designed and manufactured when Audio Components ran Audio Innovations. The redesigned circuit features lower gain 12AU7 (ECC82) valves to drive the 2A3’s. These changes are a significant departure from the initial design and an improvement sonically.
The First and Second Audio operate in push-pull class A1, a highly linear mode of operation. However push-pull topologies have become controversial partly due to losses caused by splitting the signal into opposing cycles. Paraphase phase splitters are used for this task but have a less symmetrical output than other splitter types although they perform well. Phase symmetry would become a critical issue with the popularisation of single-ended amplifiers like the Ongaku in the 90’s that completely avoid the process.
Despite the low continuous RMS output, these amps have very generous dynamic power, perhaps more so than non-triode valve amps. This yields an output power similar to transistor amps of around three times their rated output! Although more tolerant of speaker loads than equivalent single-ended amps, care should be taken with matching. Difficult or insensitive speaker loads compromise performance.
Two amplifier gain stages drive the 6B4G’s/2A3’s resulting in very high input sensitivity since there is no gain loss from using feedback. The ECC88 versions of the amplifiers have a sensitivity in the region of 100 mV! The later ECC82 version has a sensitivity of over 200 mV – high but far more usable. With such sensitivities most of the gain produced by active pre-amps is redundant. For this reason the amps, with a suitably high input impedance of 220 Kohms, were designed to suit standard passive pre-amps.
Whilst aged triode amps like the Western Electric 300B cinema amps were highly influential, The First and Second still represent the first experience of triodes for many designers, reviewers and consumers alike. Reviews at the time testify to a very different sonic experience quite unlike anything to be found in competing amps. This was the case until the early 90’s when a number of new triode amps appeared.
While current amplifier technologies offer different pros and cons, it does seem that triode amplifiers offer the kind of performance other amplifiers rarely hint at. In a way these amplifiers transcend issues like treble, bass, and detail. Although these qualities are important, the difference between an excellent triode amplifier and more conventional designs is the difference between experiencing an almost metaphysical “thereness” and the audible distance experienced with conventional reproduction. This was quite a revelation for some reviewers. There is an uninhibited intensity that the amplifiers either bring to music or do not impede, often in contrast with other amplifiers that slightly restrict the sound.
Since the 1990’s the triode has become extremely popular for high-end amplification. The higher powered 300B has become the valve of choice. More recent trends in design vary with that of the First Audio. Class A push-pull triode configurations have largely given way to single-ended triodes (SET). Although some still prefer push-pull, the majority have a preference for single-ended topologies today.
While other amplifiers of the era like the Ongaku and Pioneer A400 are better remembered, and it can be argued that the common design of triode amplification has moved on to a certain extent, it is still the case that the First and Second Audio represented a very prescient insight of some radical paradigmatic changes to come. They also stand up when evaluated as products in their own right. They are extremely fine quality amplifiers that compare favourably with many of today’s high-end equivalents.
Audio Innovations: The Story
Audio Innovations was founded by Peter Qvortrup and Erik Andersson in 1984. Their first products were the 800 Series power amp and matching pre.
The valve amplifiers of the 1980’s occupied the high-end almost exclusively. Most of the valve amplifiers of the 1950’s and 60’s were not nearly as expensive as their newer equivalents. In contrast to many other companies, Audio Innovations continued the noble tradition of making the valve amplifier an affordable item – the aim was to have one in every home!
In 1986 Audio Innovations were seemingly the first to reintroduce integrated valve amplifiers! The Series 500, a mid-priced 25 watt integrated, become probably the biggest selling valve amp of the late 1980’s while the 300 integrated was the least expensive valve amplifier then available. In keeping with the firm’s cost conscious approach early models like the Series 800, 500 and 300 cost the equivalent figure in Sterling when first introduced!
Most frequently Audio Innovations used EL-34 power pentodes. Their smaller amplifiers use EL-84 output pentodes and the Series 300 uses ECL-86 triode-driver/output-pentodes. These designs typically use paraphase phase-splitters, and Class A push-pull output stages with ultra linear feedback configurations. Relatively little negative feedback is used. Their amps were typically designed to suit conventional efficiency loudspeakers. A few rarer zero feedback pentode models like the Series 200, 400, and 800 Mk2 are extremely intolerant of speaker loads which limit their application unduly.
Erik Andersson returned to Sweden in 1986 but designers like Guy Adams, and others like Guy Sergeant, maintained the design ethos of the company. A firm called Audio Components bought Audio Innovations in 1991. While other companies were beginning to introduce exotic high-end triode amplifiers, and a prototype “Third Audio” using the then extremely novel 845 bright-emitter triode in push-pull was shown at hi-fi shows in 1990 (never released commercially), the new owners changed their approach by focusing more on the lower end of the valve market. Audio Innovations branched out into kits and even introduced a transistor amp with an old style quasi-complimentary output stage!
In 1996 Audio Partnership bought the brand. This was regrettable as they also bought a number of other once excellent firms like TDL that were then used to brand low grade products. The production of Valve amps ended, and the Audio Innovations brand was used for inexpensive lifestyle systems, shelving and silver tinned wiring sold principally by the Richer Sounds budget retail chain in the UK.
(Dagogo wishes to thank Dimitri van Hoven for providing all product brochure images in this article. -Ed.)
A note about the author:
Rob Harris has a particular interest in the historic development of high-fidelity and audio. His interests outside of audio include philosophy, music, and history/archaeology. He contributes articles to a political magazine and several websites. Rob lives in the Republic of Ireland, and he will be contributing a series of commentaries to Dagogo. – Editor