Quote:
Originally Posted by dieseljeep
A poor output transformer design is probably responsible for much of the distortion in many designs.
I know an amp designer that uses 12AT7's and even 6BK7's in the voltage amp and invertor stages.
I'll have to see if I have a Sams on the various S-C amplifiers.
|
I would agree up to a point; at the extremes of frequency the OPT must be the limiting factor. Mid-band however the output stage itself is probably the dominant term in the total distortion of the amplifier, followed by the input stage. A cathodyne operates ~ unity gain with a tremendous amount of feedback, so it's term can be dropped as negligible.
Now, that said, look at the plate curves of a 12AT7 and compare them to a 6F5, 6SL7, 5751, or 12AX7. They are all substantially more linear triodes than the 12AT7.
A 12AT7 would not be my first choice in a voltage amp. I think it is popular only because a) 12DW7s aren't cheap and b) it had better drive capabilities than a 12AX7, 6SL7, or 5751 when the second section is used as a cathodyne.
Another, superior option would be a 6AV6 with the diodes tied to ground and a 6C4 running as the cathodyne. Lots of ways to set up the voltage amp up front that would work.
My absolute favorite circuit for an input stage is the Van Scoyoc cross-coupled inverter. Gain is high, easy to wrap a feedback loop around it without concerns for stability at the high and low end, it has pretty decent drive ability all things considered, and the balance is only bested by a long tailed pair with a CCS in the tail, or a cathodyne. Balance is typically held to a factor of 1/mu+1 for the final, high gain triodes. For a 6SL7, that's about 1.4 percent, for a 12AX7 about 0.99 percent.