Hi End Tube Project

[Bandersnatch]

Originally posted by Jeb-D.


Why are you using an output stage to drive another output stage?

'cause it works well.

I'll second the idea to rid yourself of the step-down TX. A 1+1:1+1 IT with enough primary inductance to load a 6V6 pentode stage would be quite nice here. For that matter 6AQ5 or EL84 would do just as well. The 12AU7 is not worthy of designing around. 12BH7 at minimum, and there are many others better... Phase split with a LTP twin triode( or perpahs pentodes, 6AU6 maybe? ).

Also, mu of the 6C33 is more like 3. Cathode bias leaves need for a B+ supply to cover bias voltage, and plate-cathode voltage which is good for the driver stage too.
cheers,
Douglas

[Stee]

see this version two
the problem is found transformers on web

[ray_moth]

Stee,

Forgive me if I've missed something but your schema doesn't seem to make sense. What is that ECC82 supposed to be doing? The right half of it is unconnected to any form of input. That in turn means that the lower 6V6 is undriven.

[GordonW]

Originally posted by ray_moth
Stee,

Forgive me if I've missed something but your schema doesn't seem to make sense. What is that ECC82 supposed to be doing? The right half of it is unconnected to any form of input. That in turn means that the lower 6V6 is undriven.

Actually, it's a self-adjusting paraphase, IIRC.

When a signal goes through the top tube, it'll be voltage-divided between the 470K resistor (between the top tube input line and the line to the second grid of the 12AU7) and the resistor from there to ground. This will induce a voltage on the grid, which will drive the second half of the 12AU7... causing output to the lower 6V6.

However- since there's also a resistor from the bottom 6V6 input BACK to that same node (going to the grid of the second 12AU7 stage), it'll act to CANCEL OUT a lot of the gain. In fact, the second 12AU7 will act almost like an op amp at unity gain... the gain of the second half of of the 12AU7 will be set to right at -1... in theory, it'll be in perfect gain balance with the top side, just in opposite polarity...

IIRC, H.H. Scott used this topology on some of their amps, back in the old days. Seemed to work well for them, AFAIK...

Regards,
Gordon.

[ray_moth]

Thanks, Gordon, it must be my failing eyesight I didn't notice there was a resistor between the right-hand grid of the ECC82 and ground - it looked like a grounded grid.

[astouffer]

Stee: the first tube is a dual triode but it is labeled 6AU6 which is a pentode.

[Stee]

bad reading about Bandersnatch's comments

[ErikdeBest]

Hi Stee

Though the 12BH7 is similar to the ECC82, you can not just substitute them and keep the same values for the other components: the gain of both tubes is different and using same components will cause imbalances in the phase splitter.

I see you still have the global feedback loop which does not inspire very confidence due to possible oscillations.

As you already decided for a heavy and expensive IT on the EL84's, why you just don't go with another simples IT in front of that? I am thinking along the lines of the Karna
http://www.nutshellhifi.com/triode2.html

You could probably also get away with the first stage if you use triode strapped D3a's feeding into the IT (instead of the EL84).

[Bandersnatch]

Originally posted by Stee
bad reading about Bandersnatch's comments

hey-Hey!!!,
Try a long-tail-pair for the input stage. It will need a small negative supply and a CCS, but it will allow you to switch in any pin-compatible tube without wrecking the OP. It will also have very nice PSRR, though it isn't likely that you'll deliver much ripple to that stage.
cheers,
Douglas