12ax7 long tail pair phase inverter driving el84's

[nandrewjackson]

I haven't built a tube power section yet, but plan on building a (2) or (4) el84 push pull poweramp soon.
I've been studying schematics such as Vox style el84 and the Hoffman 18Watt Stout schematics.

I have a couple "theory" questions.  I understand a basic triode gain stage, and I think I see what's going on in push pull el84's, but I just don't understand how the 12ax7 phase inverter is working.

I guess I don't need to understand it to build it,    , but I would like to.   

It's quite obvious that the first grid of the phase splitter is getting the "input" signal, but how in the world does the second triode of the phase splitter(inverter) get it's signal?

I've commonly seen that the grids are either 470K or 1M each with an additional 47K above ground, with the cathodes tied together about 1K plus the 47K to ground.  And there's a cap (.o1 or .02) across the second grid R either to ground or above the 47K to ground.

Can someone give me a short "layman's" explanation of how the second grid of the longtail pair gets input?
And my logic tells me the 2 triodes would be in phase because the grids are tied together????


Thanks for your help, This Forum is THE BEST 

[PRR]

> the grids are tied together????

No they are not.

> grid ...is getting the "input" signal

Input signal of a tube is *grid to cathode*.

When cathode is pretty groundy, we might short-think "grid input"; but here we must see a bigger picture.

[nandrewjackson]

Big OOOPS

I meant to write CATHODES tied together.   



The first triode is getting it's signal with [470K plus 47K] grid to ground, and about [1K plus 47K] cathode to ground.  I get that.  That part totally makes sense.  It looks like standard triode guitar gain stage with high grid/ground R and low cathode/ground R.



I see that triode 2 is getting signal, it just seems that it's being ran in a grounded-grid fashion with more +V on the cathode than the grid.....  Is this right? 

I realize grounded-grid is not the right term for it, but would it be ok to compare the second half of the Phase Inverter/ Phase splitter to a grounded-grid triode?

[nandrewjackson]

Or instead of comparing triode 2 of the phase splitter to a grounded grid triode, should I just let the sleeping dog lie, and

faaaagedddaboudit'

 

[sluckey]

Here's a good short read about a LTP PI...

http://www.aikenamps.com/LongTailPair.htm

[kagliostro]

And here further info

DC Coupled Long-Tailed Pair

http://www.valvewizard.co.uk/dcltp.html

AC Copuled Long-Tailed Pair

http://www.valvewizard.co.uk/acltp.html

K

[PRR]

> seems that it's being ran in a grounded-grid fashion

Exactly right.

But does it help? No other audio amp runs grounded-grid, so we don't have much gut-level experience or intuition.

It may be useful to think input is "grid to cathode". Or in this case, the first cathode is not very groundy (47K away from ground) so the input is "grid to cathode to cathode to grid". Two grid-cathode nodes split the input. Ignoring the tail, they split input equally. (Tail current diverts a little signal and causes 5%-20% un-balance, which may be NBD.)

[nandrewjackson]

Thanks cagliostro and sluckey for the links, nice stuff there.
Thanks PRR for the "input" *pun inttended*

I have a better idea of what's going on there, it's just weird to see one, let alone, two triodes 47K above ground, plus all the extras, let alone if there's NFB or Presence control going on!

Thanks again,
you all are awesome!