Roscoe Primrose wrote:
Whether the grid is kept at -5V or the cathode at +5V makes no difference. The difference comes in how that voltage is achieved. In fixed bias, the difference between the cathode & grid is fixed (at DC), hence the name. It makes no difference how hard the stage is driven, the DC difference between the grid and cathode remains constant. In a self-biased circuit, the bias is achieved via a voltage drop in a (usually bypassed) cathode resistor. During class A operation, the average current through the cathode resistor remains constant, so the DC bias remains constant. When a PP stage is pushed into class AB, the average net current in both tubes increases since the tube that's drive with an increasingly positive grid voltage continues to increase it's current flow, but the tube that is cutoff by the negative grid voltage can't have the current go less than zero. When this happens, the bias voltage is increased.
Roscoe
The fixed bias amplifier, with capacitor coupling to the control grid is more prone to coupling capacitor blocking than the cathode bias (self-bias) amp. This occurs when the grid becomes positive with respect to the cathode and the grid draws current. When this occurs, the cap discharges and the amp distorts badly until the grid/cathode voltage drops to a negative value again. The cap then recharges, but the distortion is present for up to seconds. A single large transient can cause this. It happens all the time when listening to music.
In a self-biased amp, as Roscoe pointed out, once in the Class B portion of the operating range, the cathode voltage rises along with the grid voltage, making blocking practically a non issue.
Many a reviewer has marveled that a cathode biased amp sounds more powerful than a fixed bias amp which actually produces more MEASURED power. Blocking distortion is the reason why. We've now heard several fixed bias amps without output tube cap coupling and they do sound like they are way more powerful than the numbers indicate.
On the flip side, a self-biased amp is more prone to crossover notch distortion, at least in the amps I've had experience with.
If one builds a fixed bias amp, avoid cap coupling to the grid. Use a follower or transformer coupling.
Stuart