Alternative g3 suppressor grid connections?

[kilowhat]

Happy new, everyone! Been meaning to use an OT i have, that has unused UL taps on the primary. This got me pondering possible connections.
If g2 screen tied to plate approximate a triode (btw shouldnt g3 be tied in as well, for closest approx? i digress), then g3 suppressor tied to plate (instead of g2) would reasonably approximate a tetrode.

Have never heard/played a tetrode but the kinks at lower plate volts usually look extremely non-linear (see: https://en.wikipedia.org/wiki/Tetrode#/media/File:RCA_24-A_Screen_Grid_tube_anode_characteristics.jpg), which ofcourse could sound/feel like shit. Or since the kinks appear predominately at lower plate volts, im guessing theyd be a more prominent part of the sound signature under overdrive than normal conditions, as well as with higher loads and or limited damping (if reactive load). Which atleast possibly could sound interesting? Any experience or thoughts on tetrodes in guitar amps?

Onwards, if g2 screen tied to say 50% UL taps approximate a half breed pentode-triode, then wouldnt g3 suppressor tied to UL taps (instead of g2) approximate a 50-50 pentode-tetrode mongrel? Anyone tried this?
In my minds eye this looks particularly interesting since the g3 voltage would be approaching cathode/ground the lower the plate voltage gets, thus reasonably tempering the wild kinks when they are at their wildest, and presumably need tempering the most? I.e. ~half tetrode at small plate swings, ~mostly pentode at big plate swings?

Furthermore, what if both g2 and g3 were tied to UL taps? This would atleast improve speaker damping which would reasonably further temper the tetrode kinks at their most extreme.

Are any of these connections possibly hazardous? And if experimenting with this, would one need g3 stoppers for tube protection; if so what range would you roughly reckon for EL34s??

The two g3-to-UL scenarios above are of primary interest to myself, but i can think of atleast two more connections possibly viable:
<sub>1. What if g3 would be tied to normal g2 supply instead (UL taps left alone), what would this half breed topography resemble? I suppose a sort of tetrode kink tempering as in the g3-to-UL connection proposed above, but instead of directly dependent on plate voltage, now only indirectly, and more precisely correlated to g2 current, thus more specifically tempering tetrode kinks at overdrive conditions rather than normal operation plate swing. I.e. ~half tetrode at normal operation, ~mostly pentode at overdrive?
On second thought, lower screen voltage would force plate down too, so perhaps ~half tetrode at normal operation, ~mostly triode at overdrive?
2. Lastly, another (perhaps even less interesting?) possibility is g2 to UL taps and g3 to plate, which i guess would make a half breed tetrode-triode approximation.</sub>

Hope this wasnt to scatter-brainy to properly follow. Any thoughts of yours are greatly appreciated!

Many thank /kw

[PRR]

G3 design seems to be totally empirical, guided by past experience with similar-goal tubes (and minimum production cost). The effect of G3 at non-intended voltages may be small or large.

I think you don't want anything like "tetrode". There are reasons there are NO "True Tetrodes" in audio (below the 1,500V zone). '24 is about the only True Tetrode ever sold for 250V use, only for RF/IF uses with very small signals (plate never swing to 100V G2), and even then VERY quickly faded as true pentodes came on the market.

Or if you do like "Tetrode Sound", there is some way-out fuzzz pedal can do that kind of sound without busting your bottles.

Try it. Expect to try many changes and maybe burn some bottles.

[kilowhat]

PRR. thanks for your reply, very helpful!
I were sort of suspecting that 100% tetrode sound wouldnt be all . (tend to think non-linearity can sound great in moderation, but wonky in excess), hence my initial attraction to the g3-to-UL idea and a possibility to moderate the tetrode kinks.

Think ill jump ship on this for now due to too little time (have to keep reminding me this is just a hobby!) as well as limited budget to burn glass. But if i ever come across an OT with perhaps 10% rather than 50% UL tap ill try, and awaken this thread from the dead. For now, better to focus on finishing the project!

If anyone have or will experiment, feel free to hijack and post for future reference. Have tried googling the subject but no dice, except for some Traynors (e.g YGL-3A) with diametrically opposing approach, ergo tying g3 to neg bias. But as i understand it this makes the stage more prone to oscillation.

Also found this:
"...and if there is any impedance in the cathode circuit (such as an unbypassed cathode bias resistor) then connecting the suppressor to ground will dramatically increase the harmonic distortion in the anode current.* Obviously this is very undesirable for hifi designs, but might be useful for musical effects.
*Shepard, W. G., (1953). Suppressor Grid Frequency Doubler. Electronics, October p200."
Its a quote from merlins book, meta-quoted by robinette on the tdpri-forum. Since i plan on going unbyppassed i might try grounding g3 for giggles.

[PRR]

> as i understand it this makes the stage more prone to oscillation.

I don't understand.

> quote from merlins book, meta-quoted by robinette

Let's get the original up. Attached, from
https://www.americanradiohistory.com/Archive-Electronics/50s/Electronics-1953-10.pdf

The note is *specific* to the 6SA6/5725. This is not a normal pentode. G3 is designed for much more gain than any normal pentode's G3. Note also that for low Vg3, Ig2 is as large as the Ip for high Vg3, whereas most pentodes G2 current will not rise this high. (FWIW, those words are not in this article.)

[shooter]

so is your treasure-trove filed using file using the Dewey -decimal system 
nice read!