by erichayes » Wed Dec 06, 2006 8:05 pm
Hi All,
I think you're making it more complicated than it really is, Don. In class A, the tube is (theoretically) drawing the same current regardless of signal applied. In class AB and B, the current draw is proportional to input signal--the greater the signal voltage, the higher the current draw. Depending on the bias point, one or the other tube will be conducting more through a half cycle, except in class B, when one conducts and the other is cut off.
In one of the other threads, Shannon (and I) talk about running KT88s at over 60 mA at idling. If you were to look at the cathode current at that bias setting, you'd see it stay the same for a longer period as you increase the input level. That's because we're forcing them to operate closer to class A than they would be operating at, say, 50 mA quiescent. Conversely, if you were to set the bias so that, at full output (clipping), you just got rid of the crossover notch, you'd be operating in class B. When you subsequently remove signal, you'll find the quiescent cathode current is now significantly lower because the tubes aren't being called on to "work" with no signal applied.
What your meter is measuring, regardless of operating class, is DC, because the two halves of the AC waveform cancel each other out before the meter can resolve the AC component. If you were to lower the oscillator frequency to something subsonc, you'd see the needle vbrate or the dsplay to roll up and down. But if you switch your meter to AC volts, you'll find 0 VAC at idling increasing to some maximum value at full output. Better yet, hook up a 'scope to the cathode and switch between AC and DC coupling. That will show the effect on both the AC and DC components on the cathode resistor as the signal is varied.
Eric in the Jefferson State