PT VA

[Wombaticus]

Both transformers could be essentially the same size and weight. The wire used in the 350-0-350 secondary can be smaller because the duty cycle is 50% of the duty cycle in the 0-350V secondary for our case of equal power requirements from the power supply.
Tom

This doesn't sound right to me. Per my post above, regardless of duty cycle, the rms current in *each* of the centre-tapped secondary legs is NOT 50% of what is found in an equivalent non-centre tapped PT. Rather, it is 1/sqrt(2), or about 71%. So there is just no getting around the fact that the total current that the PT must handle is actually ~41% higher. Moreover, BOTH legs are physically contiguous and contribute to core heating, which is ultimately the critical issue, no?. This is not like a class B tube situation where one physically separated tube can cool as another distant tube does the work on the half cycle. It seems like making the wire diameter smaller can therefore only exacerbate the thermal issues. There is a reason why industry (broadly speaking) ditched centre tapped transformers pretty much as soon as there were good alternatives to tube rectifiers with their bothersome common cathode. Really the only major downside to a bridge recto setup is that there are two diode drops instead of one, but this obviously is only an issue with respect to efficiency for supplies intended for quite low voltages.

[Gregg]

Both transformers could be essentially the same size and weight.

I have yet to see one such PT. Also smaller size wire means higher DCR which results in higher voltage drop when tubes are pulling current.
Even an undersized poorly designed CT PTs (which as a bonus gets hot like hell) is bigger than a properly designed non CT PT.