"David Nebenzahl"
Phil Allison spake thus:

Also, the power output will of course drop in triode mode and each EL84
will dissipate more heat under overdrive. Combined with their use of a
"concertina" phase splitter, I suspect that overdrive clipping will not
be symmetrical.

If you wouldn't mind indulging my relative ignorance for a minute, that
"concertina" stage business got my curiosity up. So I found the two triode
(12AX7) stages before the EL84s, and it looks to me like the first stage
(V3A) is a phase inverter,


** With an imagination like that - you should be writing fiction for a
living.

A " phase inverter" ( or phase splitter) has equal and opposite polarity
outputs.


with the other triode acting as *another* phase inverter cascaded from the
first stage, presumably with unity gain. Is that correct?

** Nope.

V3A is a voltage amplifier ( gain of about 50 before NFB) while V3B is the
"concertina" phase inverter. Concertina circuits always have equal plate and
cathode resistors and in operation, the voltage between the plate and
cathode grows and falls with audio drive like a concertina being pumped in
and out.

The cathode of V3A acts as a negative ( opposite phase) input for feedback
from the output transformer secondary.

Because of the easy drive requirements of an EL84, such a concertina circuit
is able to send them deep into overdrive - while maintaining good symmetry.
Not so for 6L6GC or EL34 stages which need several times more drive voltage.


..... Phil