A very odd phase inverter- comments?

[GordonW]

I've got a couple of Stromberg-Carlson AU-35 amps here to work on. Originally, these were 4-input PA mixer/amps, with dual-channel output (twin PP 6L6/5881 output channels).

I'm thinking of converting one to a home-audio integrated amp. Easy enough to convert the preamp section... but this amp section is pretty unique. It's like a modified paraphase, with a REALLY odd "feedback" scheme. The part with the plate of the second section of the 6SC7 input tube being directly coupled BACK to the input (feedback injection?) is the strangest part.

Anyone got a quick summary of what EXACTLY is the operational mode here, as in what each part is doing in the signal path (schematic attached)?

I'm thinking that for a hi-fi application, it may be better to convert this over to a more "conventional" Williamson (maybe modified with output plate to diff amp feedback) or Mullard circuit... but I would like to hear input on any potential advantages (if they exist) for the current topology before I do that...

Thanks!

Regards,
Gordon.

[Yvesm]

Can't see nothing more than the usual paraphase ... curiously drawn

Yves.

[GordonW]

Originally posted by Yvesm
Can't see nothing more than the usual paraphase ... curiously drawn

Yves.

If it wasn't for that second section of the first dual-triode (the one with the plate connected into the direct signal path), I'd agree with you. That looks to be applying feedback using a very STRANGE methodology... and only to the "bottom" side of the paraphase (driging the bottom 6L6G output tube on each channel)...

Regards,
Gordon.

[Merlinb]

It is a little odd, I agree!
It's a basic paraphase, with some negative feedback (well, feedforward would be a fair description) to the grid of the 2nd 6N7 driver triode, to reduce it's gain. (Because the 220k/220k potential divider feeding the 2nd 6SC7 triode has too much gain on its own, and would make it unbalanced).

The cap from the 6SC7 anode to the input is actually positive feedback. Maybe to try and compensate for the input Miller capacitance??

[SY]

the one with the plate connected into the direct signal path

Are you referring to (what looks like) C18? That's just there to extend the bandwidth.

[Yvesm]

The summing resistors seem to be R36 and R37.

R36 from its own plate (V5?), R37 from the "normal" preamplified signal.
R39 is a common grid leak resistor for V5 and both section of V7.

In short, the signals at the grids of V7 are summed into R39 and the result is sent back to the grid of V5.

I second SY for the role of C18.

Yves.

[GordonW]

Ah, I see it now. I was visualizing the signal flow wrong. I thought that C18 was a main signal path... instead, it's just a "feedback speed-up" cap.

Didn't pay enough attention to the voltage divider just before the grid of the first half of the 6N7. That was my error. THAT'S the main signal path for the inverted phase section...BACK to the second half of the 6SC7...

Thanks, folks. Sometimes, when stuff is drawn oddly like this, another pair or two of eyes is needed...

I think I'm going to try firing up the amp stage (without the mixer- go directly into the grid of the 6SC7) and see how it performs. If I don't like it, I've got LOTS of ways I can remake the circuit...

Regards,
Gordon.