Triode question

[Earl]

In the typical 'gain stage' configureation, 100K plate load/1500 Rk and a 12AX7 - Is it possiblt to shut the tube's conduction off completely, or will there always be some small amount of plate current regardless of how negative the grid is?

If so, why is this, and is the value predictable?

[MWJB]

Yes it's possible. Though, with those values not very common.

It could happen with very low plate voltages (sometimes under 120vdc?), and/or very large grid loads (effectively rebiasing the tube). When hit with a big signal the amp can hesitate, signal can die off momentarily/notes don't always come out in a fast run.

More likely with harp (especially with crystal & ceramic mics plugged into amps with no input grid load resistor...like a stock Kalamazoo Model 1/2) than with a plucked string instrument/pickup that sets input Z to K's rather than tens/hundreds of Meg.

More likely with smaller cathode resistor values, so tweaking here and/or plate voltage can eliminate it.

Ear testing is the only way to establish the value that sets it off, in borderline cases other factors such as the instrument in Q, current bottlenecks in a power supply (e.g. loads of amps plugged into the same 4 way) can exacerbate it.

[Earl]

Thnaks for the reply, MW. I should have been a little more specific about what I'm after. I thought I might gleen the answer with a somewhat related question....

I was looking at the Phase Inverter of the basic Princeton, the paraphase style, and trying to predict the maximun plate voltage swing before the circuit clips. What I found was that, in the schematical example of the BF Princeton, the static state of the PI indicates "65V" at the cathode. My assumption was that the 180 degree swing there would be from 65V to 0V, provided enough grid signal.

What I found was (via PSpice) 65 to ~11V....using a 1KhZ sinewave input. I did include the 220K ohm (for the 6V6's, not in the circuit) bias supply resistors in parallel with the plate and cathode resistors as well as the couplers.

This circuit is unbypassed, and it dawned on me this afternoon that the 'self-bias' and resulting NFB was what was causing the 11V limit. I also remembered that there was a mod that made that stage fixed bias with a corresponding increase in output signal swing.

So then, I guess my question would be - what is the relationship between the NFB, and the missing 11V?