[DaveC113]

Professionals have used XLR balanced cabling since the 1930's because no amount of hum is tolerable in a professional context.

XLR plugs were invented in the 50's afaik. And it won't necessarily prevent hum.

[Lynn Olson]

Huh. That's not what I see in the published Aikido schematic. I see a straightforward SRPP, which is direct-coupled to a conventional cathode follower with a current source pulldown. The power supply noise cancellation is achieved via a pair of balance resistors in the middle part of the circuit.

SRPPs are quasi balanced in the sense the upper and lower input tubes have a see-saw action, with one tube increasing its current while the other decreases in complementary fashion. This complementary action cancels much of the distortion of a single tube, although somewhat differently than conventional push-pull or balanced circuits.

The gain structure is also a giveaway to the internal architecture. The mu of a 6SN7 is typically 20, and a gain of 20 can be realized if the load approaches infinity, as with a current source plate load. The gain of an Aikido is 10, which is what happens with SRPP circuits. The load for the lower input tube is equal to the Rp (plate impedance), which delivers a net gain of 1/2 mu of the tube.

[Lynn Olson]

Balanced circuitry dates from the earliest days of the telephone network, long predating the invention of vacuum tubes. Bell Labs took the barely functioning Lee DeForest tube, increased the vacuum, and worked out the math to bias the tube for maximum linearity. The Bell System/Western Electric then extended phone system methodology to broadcast and movie-sound applications.

The Raven preamp is simply a Bell System repeater amplifier, as used in the late Twenties through the mid Thirties. Modern transformers with much wider bandwidths and lower distortion are applied to a 90-year-old zero feedback triode circuit.

The famous RCA connector was used to connect a crystal cartridge in a phonograph to table radios back in the late Forties. Crystal cartridges had high output and didn't need equalization, and were fed into the first gain stage after the AM detector.

For better or worse, it became an industry standard by the early Fifties, and has been ever since.

The XLR connector, pioneered by Cannon, replaced the even bigger Western Electric connector used in the movie-sound industry back in the 1930's.

[Cloud.sessions]

I really appreciate the level of engagement this thread has generated, and as always it’s great to see both Lynn and Don contributing here.

To clarify the balanced vs. single-ended discussion: I don’t think anyone is arguing that balanced is inherently inferior to single-ended, or vice versa. Any engineer understands that both approaches involve tradeoffs, and that those tradeoffs are highly circuit-dependent. Where my earlier comments were coming from is the fairly common assumption among audiophiles that “balanced is automatically superior.” Unfortunately, the reality is far more nuanced than that.

Both topologies have strengths and weaknesses, and neither is universally better. My goal was simply to challenge the prevailing dogma, and judging by the discussion so far, that seems to be working  . The technical points I made earlier weren’t opinions; they’re established facts. That said, facts don’t exist in a vacuum — context and implementation matter, and circuit design determines how (or whether) those facts apply in practice.

@DaveC113 — great to see your contribution here, and you’re absolutely right regarding the Aikido circuit.

@Lynn Olson — as always, a master class. I wouldn’t describe the Aikido as a conventional SRPP so much as a variant or close sibling rather than a textbook implementation, but I think we’re largely debating semantics at that point.

As for the Raven and The Dawn: this thread isn’t intended to be a full design overview. I’ll start a separate thread as we get closer to release. That said, I do want to clear up some confusion.

The Dawn is designed to be a more accessible preamp, with its lineage rooted in the DS2 (Aikido) and implemented as a single-ended line stage. We didn’t feel we could bring the Raven to market at a significantly lower cost while preserving the level of performance we consider essential — part of that is due to the Raven’s balanced architecture, which needs to remain balanced for the circuit to operate optimally.

That said, The Dawn is a very different animal from the Raven. Because the designs diverge so fundamentally, each circuit has its own inherent strengths and limitations. Our approach with The Dawn was to lean hard into those strengths while minimizing the drawbacks. The design goals were an ultra-low-noise preamp with excellent timing coherency and wide bandwidth, while preserving what we believe is one of the core virtues of tubes: their harmonic character.

We don’t see even-order tube distortion as something to be “designed out.” In our view, that is the feature.

We’re confident we’ve exceeded our original goals. Current measurements show a noise floor below –115 dBV, with a textbook harmonic profile dominated almost entirely by second-order content and vanishingly low higher-order components. To put that in perspective, fourth-order distortion is buried below the analyzer’s noise floor  . I’m not aware of many commercial tube preamps under five figures achieving that level of performance.

That said, I won’t give everything away just yet — more details will come when we’re ready to formally introduce the product  .