Class A2 Operation
Class A2 Operation
“I pushed in the 6EP7 and GOD spoke” How many times have you heard this? Over and over, the driver tube of this month’s asylum reveals the latest, greatest driver tube ever. Best one was, “its better than live”. Has no one heard of transfer characteristics or operating points of tubes? A circuit with one tube type can be changed to sound like another tube by just changing its plate voltage, bias or load impedance. When a different tube is used, it’s different also. When changing operating points, many like tubes can sound exactly the same. What they hear and unknowingly argue about and question continuously is nothing more than the different dynamics of each change. More correctly stated is “they sound different, not better”. Now more than ever, the one that can argue the best is therefore right.
A standard tube power amplifier operating in class A, has low distortion and low power output. No grid current is drawn in class A. If this stage is driven harder and grid current happens, severe distortion is heard. This is from the bias, now changing with the signal the grid sees (through the grid resistor impedance). Therefore the grid bias changes and the operating point also changes. The signal becomes part of the bias. This is all this output stage can do.
In A2 operation were the grid is driven with a driver that delivers a terminated high power level to the grid, the grid has no resistor in series with the bias, the bias is capable of the current of that driver so, it can‘t move its operating point. The same tube can deliver almost 3 times the power provided its power supply can handle it. Here then are several examples. A 2A3 driven in conventional class A, delivers 3 watts of clean audio while this same tube in A2 will deliver 8 watts of clean power with about 3 watts of terminated grid drive. A 300B driven conventionally delivers 8 watts but if A2 driven it will deliver 20 watts with about 3 watts of drive. Push pull 300B’s in A2 deliver 50 watts. The A2 output tube will give, midpower/midband, less than 1% distortion, provided the plate load impedance is a little higher than the popular 3k ohms. It is important that the load be addressed in A2 operation; it is more critical than operation in A. Increasing the load to 4K+ will clean the distortion up.
What driver tube will offer 3 to 4 watts with less than 1/2% distortion of drive for A2 operation, none! You can use another amplifier, How? It can be done but this will require an interesting transformer.
The state of the art op amp is so linear few distortion analyzers measure its distortion. They also are very uniform from type to type (in audio use). They are used in the highest quality sound recording equipment, exclusively, so I see no reason to not use it for an output tube driver. “You can’t do that, that’s a solid state amp in a tube amp!” is the usual response. Then we reply, “Can you visualize a CD player that is totally built with tubes and not one solid state device in it? Why it would be the size of a refrigerator!” CD players have many IC op amps in their design. So what’s the difference - nothing. But I still want a tube to drive my speakers. Tube voltage on an output transformer sounds alive, that has been the reason SE has increased in popularity over the last decade, not the caps, interconnects or tube sockets.
Chip amps delivering up 10 watts of clean power (like TDA 2030) are easily built. A2 grid loading transformers have been designed for chips like the one mentioned, such as the A2LD-7. There are no tubes on this earth that can do what a well designed chip amp can do. The operational impedances selected from many experiments are from 20 to 30ohms for the chip to 10 to 20k ohms terminated for grid. This arrangement offers a very close conventional gain series for input sensitivity to power out. The Higher input impedance offers a more stable load for the chip and the 10k to 20k ohms provides high drive voltage with power behind it. These interstage transformers are not too costly but are intricate due to the need for wide bandwidth.
Last updated: Thursday July 5th, 2007 at 07:44:02 PM