Which is the best type of clipping in tubes?

[otto]

Which is the best type of clipping in tubes and which one gives more sustain?
I have in mind 3 types (if what I write is right):
1. the one obtained when grid goes to a negative voltage below cutoff and no current flows
2. the one obtained when grid goes to a positive voltage above cathode voltage
3. the one obtained when, with a cathode resistor unbypassed, grid and cathode voltages swing together with an amplitude such that the cathode is at zeto volts for a part of the cycle

[Enzo]

The short answer is that which you prefer the sound of.

Someone more into the math could give you better insight into your actual question, but I worry more about trying to overthink a problem.

Your input stage has a lot of gain, and you won't clip it easily with a guitar, The signal at the plate can be really large and still quite clean. The purpose of the input stage is to lift the guitar signal up above the noise floor a good ways so the rest of the amp can then work with it to the desired end.

Clipping at each stage in an amp will have a different result, a different sound.

I never thought of clipping as providing sustain, to me it was the gain of the amp that amplified the signal enough to drive the strings further. The clipping was incidental to that gain. That is an opinion though.

The amp is not homogeneous throughout. Which stage clips and how is an important question in your case. It is one thing to overdrive a gain stage early on, it is another to clip the phase splitter.

In each of your cases, where in the amp are you expecting it to happen, and what is the likelihood ot could occur there. Seems unlikely that little 200mv or so from the guitar is going to do any of that at the input stage for example, so the question, at least there, is moot. So what would cutoff voltage be in a given stage, and can the signal get there?

[otto]

mmm
I'm speking about clipping in a 2nd stage