A 2A3 PP will be a great unit in any case. The 6J5 (or its close relative
6C5) is a great tube, but it would be simpler to use 6SN7s. Personally, I
prefer "long tailed" phase splitters to cathodynes, don't know why but they
sound more "vivid". Suppose You place a 6SN7 at the entrance as a diff amp,
then another one as a LT phase splitter and driver. Using the 2A3 in PP,
with cathode bias, You'll need a 350V B+, which is high enough to get the
required voltage swing out of the long-tail pair. If You don't mind using 4
tubes, the 6J5s can be matched to minimise distortion (it's difficult to
find 2 similar triodes in the same glass envelope, therefore diff amps with
6SN7s could show some minor unbalance).
The diff amp at the 1st stage will provide a feedback entry too.
As per the output transformer, any good 3k5 unit will fit. BTW, being a PP
unit, it can be fed AC to the filaments, no problems. I'd keep separate
windings for each tube in order to have each one individually biased; this
allows to use any tube and it is safer in case of sudden failure of one
tube, but it is obviously easier and cheaper to have the 2 tubes of the PP
pair fed by a single transformer winding, having a center tap referred to
ground by the common cathode bias resistor (+capacitor). In this case, in
order to match quiescent current, tubes shall be matched on a good tester
(or bought from a dependable source as matched pairs),

Let us know..

Ciao

Fabio



"Andy Evans" ha scritto nel messaggio
oups.com...
I'm building a PP 2a3 amp right now and haven't decided on the input
stage. I'd like to use 6J5G and/or 1626 valves. Open to other
suggestions, but like to keep to octal or UX5 types (I also have 76,
37, 56, 27). I'd also need a differential first stage for the option of
balanced or SE, and I have CCSs. 2a3 needs about -50v of bias. Ideas?
ANdy