Concertina Parafeed, or How to mess with OldEurope's mind...

[kenpeter]

This final can't have a voltage gain any higher than 2, now can it?
The question here, is this considered negative feedback or not?

Suppose the output tube had been selected for a Mu of exactly 2,
would there be any negative feedback to the cathode??? 6AS7GA
for an example. Would voltage gain still be two then, or unity???

[smoking-amp]

The output stage is equivalent to a single ended circlotron. Voltage gain of two (across whole primary, for infinite Gm at V2 when loaded at secondary, or finite Gm at V2 when not loaded) due to the 50% CFB (cathode feedback) for V2. V. gain less than two for finite Gm on V2 when loaded. (This is all assuming that V1 is a small signal driver with a high value resistance between the cathodes, not a big honking power tube.)

For Mu of 2, and no loading, the V. gain would be two across the whole primary (with respect to grid1 drive). So cathodes of V1 and V2 would be V. tracking. For output loading, the second stage gain would drop though, so cathodes would not be tracking then. (hand waving calculations, perfect xfmr, perfect CCS's, constant Mu ...)

The cathode to cathode resistor could be used to enforce the gain of two better by neg. feedback to V1, if V2 is not holding up the gain of two. Essentially puts more Gm in the loop than what just V2 has available.

Don

[Merlinb]

Originally posted by smoking-amp
The cathode to cathode resistor could be used to enforce the gain of two better by neg. feedback to V1, n

Positive feedback, between the cathodes that is. Which I wouldn't trust with a barge pole...

[kenpeter]

Nah, cathode resistors are only for back coupling anti-hum for
better PSRR.... Ask OldSpice why that path ain't NFB? I forget.

Lofty Whitguy sez all three terminals of V2 float completely on
common noise, and are therefore immune to it. Pretty sure I
did nothing substantially evil to screw with that original magic.

Just throwing a new curve at an old design. Another way to
parafeed SE off a constant current...

I tried not to have any NFB except that with was plausibly
intrinsic. Or (like the CHB chain) not actually doing anything
NFB'ish taken in the full context of other signals around it.

[oldeurope]

Originally #4 posted by kenpeter
Nah, cathode resistors are only for back coupling anti-hum for
better PSRR.... Ask OldSpice why that path ain't NFB? I forget.

Lofty Whitguy sez all three terminals of V2 float completely on
common noise, and are therefore immune to it. Pretty sure I
did nothing substantially evil to screw with that original magic.

Just throwing a new curve at an old design. Another way to
parafeed SE off a constant current...

I tried not to have any NFB except that with was plausibly
intrinsic. Or (like the CHB chain) not actually doing anything
NFB'ish taken in the full context of other signals around it.

Hello kenpeter,

It is not my fault that you don't understand
the difference between the Loftin White topology
and the single ended circlotron.
Please don't write in slang and don't
write honourable names in a condescending way.

Darius

[kenpeter]

I like Oldie. His "rules" require one to think inside such a very
small box, that you actully end up thinking outside the box.

[Boris_The_Blade]

Originally posted by oldeurope

Please don't write in slang and don't
write honourable names in a condescending way.

I refer you to your 'technical forum rules'

edit: almost forgot the wink - - there we go

[kenpeter]

Originally posted by smoking-amp
The output stage is equivalent to a single ended circlotron.

I wouldn't go so far as to say your interpretation is wrong.
But a Circlotron pulls opposing DC from through the entire
tranny. I only pull DC from the center as would Push-Pull
or SEPP.

I think it is clear that the output stage of this circuit is SEPP,
based upon a Concertina Splitter, with the cathode driving
the opposing end of a PP OPT in Parafeed. It is also clearly
floating with the PS noise exactly as does a Lofin White.

Circlotron(ish) only in having the second CCS, which one
might say was a 2nd power supply. But not independant
floating power supplys, as both B+ ends are tied together
at the top.

------------------------------------------------------------------------

Darius,

Just cause I draw a clown nose on a copy of the MonaLisa,
doesn't mean the result isn't art. Its not as if I somehow
defaced the one and only priceless original. Get a grip!

I am unable to determine just how seriously you are taking
this whole Loftin White topology as a religeon? If we are not
both having equal fun in these discussions, I'll ease up and
go work on something else for a while.

I still find your ideas challenging and worthwhile.

[kenpeter]

Originally posted by Merlinb

Positive feedback, between the cathodes that is. Which I wouldn't trust with a barge pole...

Actually, V1's cathode resistor and the "CHB" network between
both cathodes must do at least three things simultaneously:

1) Provide a DC bias for V1's cathode
2) Sum of positive and negative feedbacks = 0 at V1's cathode.
3) Inject a weak sample of power supply ripple, just enough to
float the plate of V1 and grid of V2 in common to the B+ rail.

[smoking-amp]

re: Merlinb
"Positive feedback, between the cathodes that is. Which I wouldn't trust with a barge pole..."

Yup. I missed the phase there.

re: Kenpeter
"I think it is clear that the output stage of this circuit is SEPP,
based upon a Concertina Splitter, with the cathode driving
the opposing end of a PP OPT in Parafeed. "

But SEPP has two active devices to drive a single winding in both directions. Are we looking at the same schematic here? I only see one active device driving a double winding, like in Circlotrons. (the parafeed cathode part is driving in the same phase as the tube plate part. Only the CCS drives steady DC operating current in the opposite polarity.)

re: Kenpeter
"2) Sum of positive and negative feedbacks = 0 at V1's cathode.

What you have here is called "Hawksford Error Correction". (actually invented by Llewellyn in 1941 using tube circuitry. Llewellyn, 1941, patent # 2,245,598 ) You can search on Hawksford Error Correction in either the SS or Tube forums. Mostly in the SS forum. When the output stage gain is correct, no feedback occurs. When it is wrong, error correction is sent back.

It's a positive feedback loop that can be adjusted for exactly unity gain (of error correction) around it. (it actually becomes infinite signal gain around the loop then. A breath away from oscillating.) This part of the circuit is actually quite interesting. Not normally seen in tube circuitry. Although the Wollcott amplifier comes very close.

Somehow, I don't think Darius will buy into this high positive feedback scheme though. (rule # 2584432)

Don