Triode for 4XP replacement

[Mike Watterson]

How did you connect the 6V6 as a triode?  G2 to Anode?  Any parasitic stopper resistors?
The beam plates are already internally connected to the Cathode.

The main reason to connect g3 as well as g2 to triodize a pentode with wound g3 is HF/VHF capacitance. A secondary reason is if it's a spiral wound grid. So EF80 or DF97 at VHF will have g3 and g2 tied to anode to work as a self-oscillating mixer. It's not a problem and possibly better that beam Tetrodes (Philips called their Kinkless Tetrodes aka Beam Tetrodes, pentodes) have "g3" to cathode or 0V. Because a Rod Pentode has no spiral grids, they too have g3 to cathode or 0V. They are staggeringly low capacitance compared to 1L6, DF97 or EF80, so they actually work with g3 at 0V triodised.

Elsewhere I discussed the 1p24b as a possible PX4 and similar replacement. The filament can be series and will work at 2V @ 110mA (but decouple centre tap). Or you can use ballast resistor for 4V supply to give 2.2V, or have two in parallel with filaments in series for 4V @ 110mA. You can put three in parallel and ballast resistors for 2.2V max. If the 4V is centre tapped or otherwise center referenced (like a pot to 0V), then 1/2 the ballast resistor at each end. The data sheet specifically discusses triode mode and power.

I do think 6V6 or similar triodized is a good idea if you can find 6V. The grid bias should be considered. Any KT or beam tetrode is fine with g3 to cathode when g2 is to anode. c.f. UL windings on EL84 etc.

 
Any thoughts on the 4v to 6.3v transformer? I have a winder, I would hope to make it small enough to fit in an adaptor base.
Sam.

It's 450mA. One idea might be a 4V to 2V (6V is enough) transformer and put that in series with the 4V then it only needs to manage about 0.9W rather than 2.8W, but still might not fit.