Long tailed pair schematic questions

[martyh]

Hello, I have been lurking at this forum for a few years and have generally been able to find the answers I need in the archives. I am on unfamiliar ground with this schematic and am hoping someone can help me out.

I am considering some modifications to a pair of 1568A Altecs including this modification to the input as a replacement for the existing voltage amp and cathodyne splitter.



The original output transformer has a feedback winding that is normally grounded on one side with the other side taken to the cathode of the input tube through a 1500 ohm feedback resistor. I would lift the normally grounded side and use it for the other phase of feedback shown in the schematic. Since the original load on the feedback secondary is 1500 ohms, should I be reducing the size of the two 1500 Ohm feedback resistors in my schematic to 750 Ohms?

And the biggest question of all, are there any other obvious mistakes in this schematic?

Thanks in advance for any advice you can offer.

Marty

Original schematic

[ErikdeBest]

If you are using a SE input the easier way is to connect the feedback to the unused (now grounded) grid of the rightmost triode. Much like this input stage.

[Yvesm]

Well, but why not simply inject the feddback at the free grid of the LTP ?
Remove the by pass cap, choose a convenient value for grid leak -probably in the 100 ohms range- that will form a voltage divider with the existing 1K5.
Since the open loop gain will be different, so will be the FB ratio, adjust this supposedly 100 Ohms for the expected sensitivity ...

As Erik suggested while I was typing !

Yves.

[hifimaker]

It's interesting how precious few tube amps have balanced inputs. It would be great to see some successful designs. Balanced inputs are great for reducing noise in SS, why aren't balanced inputs more prevalent in tubes?

-David

[Brit01]

Sorry just to sidetrack a little.

What software are you using to create your diagrams? I'm looking for some freeware.

Thks

[kevinkr]

Not to mention that using the proposed cathode feedback connections make the use of a CCS on the tail pointless from an AC standpoint as the pair will no longer act like a differential amplifier, but more like two separate single ended voltage amplifier stages.

[kevinkr]

Originally posted by hifimaker
It's interesting how precious few tube amps have balanced inputs. It would be great to see some successful designs. Balanced inputs are great for reducing noise in SS, why aren't balanced inputs more prevalent in tubes?

-David

There have been a number of balanced tube designs most notably from BAT, but also from ARC, Berning, and others. I briefly built amplifiers and pre-amplifiers with balanced I/O, but found it to be mostly unnecessary unless runs were relatively long. See my old 300B PP amplifier design in VTV for an example of a balanced input tube amplifier.

I'm not at all convinced either that it sounds any better than properly implemented SE circuitry and interconnects and it adds a lot of cost when done properly. TANSTAAFL applies.

[martyh]

Originally posted by Brit01
Sorry just to sidetrack a little.

What software are you using to create your diagrams? I'm looking for some freeware.

Thks

I am using AutoCad which is probably not a great choice if you are looking for a program because it is expensive and not so easy to learn.

[martyh]

Thanks for the advice, is this a better scheme? This was where I started after looking at the very schematic that Erick linked to. I could probably just use a 1k pot and a 500 ohm resistor in the as the feedback voltage divider.

Would there be any advantage to cap coupling the feedback voltage divider to the grid and using a grid resistor of the same value as the input side? Something like this


Thanks again,
Marty

[Eli Duttman]

Marty,

That's getting close, but you really don't need a cap. on the non-inverting side. Use the NFB voltage divider's bottom leg as the grid leak resistor. Eliminating that cap. reduces the possibility of phase shift induced instability.

Another thing to keep in mind is stage gain. The theoretical max. for this type of driver is mu/2. With its mu of 20, I have doubts about the 6CG7. The 12AT7, with its mu of 60, easily yields 15+X gain and sounds terrific too boot, in this application. Hint, hint! Google El Cheapo.