High Perveance-Enhanced Class A3 Tube Operation
To start off with, the 300B and other favorite audio triodes, being the tubes of choice over these many years, are possibly on their way to retirement. For example, the 300B class A operation, 400 V plate to filament, at 100 mA is 40 watts, full plate dissipation, this offers a 20-25% output power of 8-10 watts, into a 3K primary. The maximum power is at slight symmetrical clipping. At that point, you always see a rise in plate current. This rise is calmly accepted and never pursued due to the fact the “just more distortion there” plate will get red at this point anyway.
Class A3 Operation (2010)
Using the 300B example and subject of experimentation Class A3 became a practical addition to power triode tube operation using fairly inexpensive triodes, such as the Sovtek 6A3 as a 300B replacement so one did not have to seek out expensive 300Bs and risk damaging them. Operating this tube at a higher plate voltage, lower plate current, lower plate dissipation, and higher load impedance (500 V @ 70 mA plate or 35 watts plate dissipation with no red plating) gave an output of 14-15 watts to slight clipping with very acceptable distortion measurement. This shows a 40% efficiency in this new class of operation. Also, the drive needed was much less than the standard Class A used.
Enhancing Class A3
Using the Sovtek 6A3 and increasing the plate voltage to 700 V @ 60 mA (42 watts plate, a slightly higher plate dissipation and a slight red plate), this increased plate loading generated 20-22 watts of very clean power resulting in a power efficiency of 52%. Many amps have been built with this type of operation with no failures to date.
NOTE: Do not use original or new production 300B’s (including the expensive boutique tubes) as they may not survive the differences in electrical specifications which may result in flashover from gas, especially in the new old stock 300B’s.
Introducing High Perveance-Enhanced A3
Having found the limits of the basic popular SE power triode and some others such as the miniature 5687 (with an output of 7 clean watts from one tube - paralleled both sections - at a total of 24 mA with 525 V on the plates and slightly red plates), we understood red plates are not something widely accepted within the audiophile crowd. However, slightly red plates do not compromise the tube at all. For example, Eimac, the transmitting tube company since the 1940’s has been using full red plate power tubes with forced air cooling even to this date.
So, looking into other tubes such as “High Perveance tubes” (large cathodes) and power pentodes, we found an interesting set of rules. We also found which ones work and others that dangerously arc over on their sockets in this operation. The EL34, was just one example and is now useless in this type of operation. The addition of plate caps on top of tubes allows them to be used with high plate voltages without arcing. For example, the 807 tube is just a 6L6 with a plate cap and a rating of 750 V plus signal.
Experiments with these tubes showed real promise and very interesting features. We did the usual experiments and found we could get near 100% power output of the tubes rated plate dissipation without red plating. An 18 watts plate pentode gave a very good delivery of 17 watts output with no red plate. The screen added to this efficiency. Using a tap of various percentages gave us a look into the heart of these tubes and showed the need for various ratios of tap percentages, like the UL tap on push-pull outputs. This tap also adjusts the distortion figure due to zero negative feedback being used in this circuit.
This tap is in phase with the plate, so it speeds up the electron flow, thus more power. Paralleling these tubes gave the same results as the standard operation and half the load impedance. Up to 38 watts resulted in one test. The one feature that caught our eye was the very low idle current used, resulting in an output transformer 25% of the usual core mass needed to achieve the same power and full bandwidth as in a standard Single Ended Class A operation.
Theory of High Perveance Power Pentodes Characteristics
Having a full library of vacuum tubes back to the 1920’s and building my own tube transmitters in the 1950’s, to date, this new operation of power pentodes is not in those books at all, not even a hint. We then looked elsewhere only to find these interesting actions in books on Particle Physics and the help from a very good friend in the particle physics field, Dr. Rob Ashlock. The theory holds fairly true of these characteristics we saw.
First, electrons accelerated at higher speeds with higher voltages will produce more power colliding with its positive destination (the plate). Second, electron clouds forming around the emitter (cathode) are drawn through a control grid at a given speed to its positive element based on the vacuum and voltage potential, this is normal Class A operation. Third, the screen acts as an electron gun accelerator as in a television picture tube and in-phase, signal wise, offers the effect of velocity modulation increasing power gain on signal peaks. This is based on the doubling of electron voltage increasing the power by a factor of four.
Thus, the result is a higher plate voltage, higher screen voltage accelerator, and lower electron cloud density (lower cathode current) with the amplitude of the signal being velocity modulated instead of just a constant current controlled gate, as in standard class A. This offers a 70-80% increase of power output efficiency over the old Class A Triodes.