I should mention the Raven preamp was partly influenced by borrowing a Rowland Research linestage preamp to use with my Amity PP300B amplifier I designed around 1997 or so. The Rowland was the only solid-state preamp that was tolerable with the Amity, which was quite unforgiving of input sources.
I was curious why it didn't sound "solid-state" at all, unlike every other SS preamp I had borrowed. True, it's built on a solid NC-milled aluminum chassis, which gives it a classy (and very expensive) aerospace air, but solid-state has quite low sensitivity to vibration, with the exception of capacitors (which are somewhat microphonic under certain conditions). What was more interesting was a Jensen JT-10K input transformer driving an ultrafast opamp. Ordinarily, that would make little sense, since the particular opamp Rowland used was so fast it could easily pass a video signal with no degradation, while the studio-grade transformer has a steep rolloff around 80 kHz or so.
And then it dawned on me. The Jensen pro transformer was effectively a signal conditioner. It scraped off all ultrasonics, including RF noise far outside audio band, so the opamp could never slew under any conditions. And blocked DC, of course, and also removed an astounding 80 dB of common-mode noise from the input cable.
I didn't buy the Rowland, of course, since I'm a tube guy, but it gave me food for thought. The same principles could be applied to a balanced tube stage, provided an output transformer was also used, along with an accurately balanced volume control (using precision stepped resistors). So the Raven was born, a zero-feedback balanced tube linestage using transformer coupling throughout.
I was technical editor of Vacuum Tube Valley magazine at the time, so I carried out research to see if Western Electric had gotten there seventy years before ... and they had. The Raven is pretty similar to late-Twenties Bell System telephone long-lines amplifiers, easily capable of pushing a voice signal down many miles of balanced twisted-pair wire. Well, modern computer-designed transformers are far superior than anything in the Twenties, so why not?
The inherent distortion of 5687, 7044, 7119, and 6SN7 tubes is quite low, and there's ample power to drive hundreds of feet of cable. And balanced operation linearizes the tube by about 30 to 35 dB while rejecting noise by the same amount. Give the whole thing a modern precision-regulated linear power supply (no switchers) and the only noise should be a very low level of tube hiss ... which is fully uncorrelated with the audio signal. (Signal-correlated noise is an annoying form of distortion that fuzzes up the music, while fully uncorrelated noise sounds like soft rainfall outside the house. In practice, the Raven is dead silent, with no hum, hiss, or buzz at all.)
Don was the first to take this idle thought experiment of mine and reduce it to practice, throwing in many improvements that were not possible in 1997, such as the precision regulated supply and the accurately balanced volume control, complete with a remote control with a L/R balance feature as well as volume control. So the Raven is a fascinating mix of 1920's and 2020's technology, sweeping completely over 1950's and 1960's design approaches.
