Paraphase confusion

[ben-horn]

Could someone explain the different Tweed Paraphase PI's to me.
So the 5C8 Twin and 5D8 Twin are both paraphase ,right ?
Schematics:
http://www.fenderholic.com/schem/twin_5c8_schem.gif

http://www.fenderholic.com/schem/twin_5d8_schem.gif

Is the 5D8 a floatig paraphase ?
are they very different soundwise ?
What's your opinion abaut paraphase vs. cathodyne ?

Thank you

[harry]

The 56k shared resistor in the 5D8 makes a "balanced" paraphase IIRC.

I'm not too sure on the theory I'm afraid; I guess it's just an attempt to offset the extra gain caused by the inverting triode gain stage.

[pdf64]

For the 5C8 and 5D8, the 2nd tride of the phase splitter is arranged to have an inverting gain of 1.
The 5C8 uses a passive potential divider at the 2nd triodes input, such that the attenuation at the input should just about match the stage gain.
The 5D8 uses negative feedback, the 2nd triode has 2 x 220k resistors to set the stage gain as 1 (inverting). The 56k is just a grid leak resistor, as there's a virtual earth at that point the value could be anything over a wide range, say 10k to 1M.
Soundwise, it's best for you to but with any reasonable arrangement, the power tubes will be saturating before the phase splitter does.
The Vox AC50 uses a take on the 5D3 arrangement.

[tubeswell]

FWIW Merlin goes into this a bit on his site

The Valve Wizard -Paraphase

The Valve Wizard

[MWJB]

"The 5D8 uses negative feedback, the 2nd triode has 2 x 220k resistors to set the stage gain as 1 (inverting). The 56k is just a grid leak resistor, as there's a virtual earth at that point the value could be anything over a wide range, say 10k to 1M." Not quite, adjusting the value of the 56k will audibly skew the balance of the PI, you'll notice that the 5D8 actually uses 220K & 270K (not 2x220K), the 56k helps balance the PI.

The 5D8 arrangement is easier to build, the value of the quoted "6.8K" in the 5C8 is a lot more sensitive to inbalance and can sometimes need careful tweaking.