6U8/ECF82 shielding question

[tubeswell]

I am contemplating using the pentode side of one of these as a V1 preamp stage, with the triode as a CF stage driving a tone stack. I have some questions about the shielding (or lack thereof compared to (say) an EF86).

The 6U8/ECF82 doesn't have a built in shield, but the datasheet http://r-type.org/pdfs/6u8.pdf mentions/implies different inter-electrode capacitances when shielding either to Pin 4 (heater filament) or Pin 6 (triode plate) or not at all.

1) I presume you would only shield to Pin 4 (heater), if you were using a DC heater supply? (anyone?)

2) If you shielded to Pin 6, the shield would have to enclosed with another shield at ground potential to be safe. (But isn't that still bouncing around under signal conditions? And wouldn't the plate resistor induce (some kind of shot?) noise coupling into the shield?)

3) I suppose you could 'shield' it with a garden variety socket shield (attached to the chassis at ground potential), but how effective would that be? Methinks there must be a reason why they don't mention it (or isn't there?), or why they suggest specific pins for the shielding.

4) What is the risk of not shielding it at all?

TIA

[jjasilli]

Also checkout the GE specs: http://www.mif.pg.gda.pl/homepages/frank/sheets/093/6/6U8A.pdf
The capicitances are very small in the single digit uuF range or smaller.  Probably won't effect human hearing, and is more important for use of the tube as an RF oscillator.  Anyway, why not tryout a SW or jumpers, and see if you get a difference.