Output: 80W / Channel; 8 Ohms
Frequency Range: (TBD)
Fixed Bias
Active Screen Voltage Regulation
Variable gNFB (Range: (TBD))
Low Noise, Solid State Power Supply
2) 6BQ7
2) 807
1) 6X8
1) 0A2
The 807 started out as the 6L6 when O. Schade of RCA developed this type in 1936. The 6L6 was designed as an audio final. However, the design proved so successful that the RF version, the 807, soon followed. The main differences being that the 807 is a glass envelope tube with the then common five pin base. The original 6L6 was a metal tube with the new Octal base, which was originally intended for the new class of metal tubes. Being an RF type, the 807 included a top cap plate connection. This feature not only simplified the construction of RF equipment (since the chassis could serve to shield the input from the output) but allowed for operation at higher voltages. The A Number One problem with the 6L6 is its "7AC" pin-out which places the plate at pin #3, and the heater connections at pins #2 and #7. The highest voltage connection is next to the lowest voltage connection. At the time, this was not considered a problem, since the 6L6 was never meant to operate at much above 350VDC or as anything other than as a Class A*1 amp.
Back in "the day", Class AB2 operation was not favored for audio amplification except where power and efficiency had priority over sonics (PA systems, AM plate modulators). The A Number One problem was delivering the required grid current from a low distortion driver. It is a formidable challenge to design a hollow state driver that can operate into the greatly varying impedance presented by control grids that swing positive. During negative half cycles, that impedance is quite high. During the positive half cycles, when the grids pull current, that impedance drops enormously. This is not a problem in RF amps since whatever harmonic distortion results is removed by LC tuners and/or bandpass filters. It's no wonder that there are relatively few Class AB2 audio designs from those days.
The usual solution to the problem of lotsawatts was to parallel up Class AB1 finals. This, of course, brings with it the problem of matching characteristics across even more VTs.
Today, the power MOSFET answers the grid drive problem. Having far higher gm's than any vacuum tube, the MOSFET source follower has a much lower output impedance. It also has the current sourcing capability since the rdon of most MOSFETs is well below the plate resistance of any VT. There is no longer any reason to avoid Class AB2 operation, and the 807 has some of the lowest Class AB2 THD specs (THD= 3.5% before any NFB is applied). Since the power MOSFET source follower has such a low output impedance, it has much less distortion than any of the commonly used hollow state drivers: transformer coupled power amps, cathode followers, or SRPPs. If you need the extra power, the Big Vix can deliver watts without sonic compromise.
Active screen voltage regulation is also included, since pentodes operate with much less distortion if the screen voltage is tightly regulated, and returned to AC ground via a Lo-Z path. The active regulator answers both these requirements. This is especially dritical for a Class AB2 design, since the screen current variations are larger than those encountered in Class A*1 operation.
The front end voltage amp / phase splitter is an LTP of cascoded 6BQ7 small signal triodes. The 6BQ7 was originally intended for use as a cascoded, VHF small signal amp. Even though audio amplification is not mentioned in the spec sheet, the 6BQ7 does offer excellent sonics, though must be run at VPK's that are higher than what would be "normal" for small signal audio triodes. The voltage problem doesn't occur with this design, since the voltage is already quite high. The front end also has generous "headroom", so that the finals will clip before any other stage. This helps to improve the over drive behavior considerably.
Cascoding offers the advantages of small signal pentodes: a greatly reduced CMiller and high voltage gain while retaining the sonic signature of triodes. This topology, since it reduces the plate voltage swing across the triodes, also improves the linearity. The extra gain eliminates the need for more than one gain stage, while allowing enough reserve gain for gNFB with decent input sensitivity. Since the 6BQ7 is designed to have a decent gm with a smaller rp, it can source the current the grid drivers require at high frequencies to avoid the slew rate problem.
To further improve the AC phase balance, and the harmonic distortion between phases, an active tail load of cascoded BJTs is included. The active tail load represents an equivalent AC impedance that's significantly greater than that attainable with passive tail loads and reasonable DC supply voltages.