Yes, I see now that you did say an elevated heater system doesn't need to be balanced. That makes things a little easier, eh? It's worth nothing that some tubes like the 12AT7 have an h-k limit of less than 100V, and it isn't always clear from the data sheets whether that might include the AC component of an unbypassed cathode. This limit requires care if the entire string will be elevated on the basis of LTP voltage. Actually, I would expect signal levels at the LTP to be high enough to forego this method of hum reduction. Just raising the voltage to suit the preamp should be enough. Thinking about it further (which usually gets me in trouble) this might be a valid reason for balancing the string, even if it's elevated. It might be that some of the tubes in the string will experience a positive heater, others not. Just thinking out loud here...zaphod_phil wrote: ↑Fri 01/21/22 6:52 pmYes, that's just what I was saying earlier. And so there's no need for a centre-tap on the PT's heater winding, whether real or virtual. However, keep in mind that the cathodes of the LPT-PI in a typical Marshall 18W are at 70V, so you may need to elevate a bit higher than that.
A handy way to get a heater-elevation bias voltage is to tap it off from the 220k filter cap bleed resistor, by splitting it into a 50k|170k 1W potential divider pair, with the 50k to ground and a 16uF cap in parallel with the 50k. That should give you 77V clean elevation voltage.