DC Heater 0V referencing

[Gingertube]

I'm about to build an AFD style clone. Parallel 12AX7 input stage with DC heater
Should I:
1) tie the 0V side of the heater DC supply to signal ground (0V) locally at the input tube
2) let the DC supply float
3) elevate the DC heater supply to say +50V
4) AC couple both sides of the heater DC supply to signal ground (0v) locally at the input tube via say 470nF poly caps.

I'll be using Mullard ECC83 pulled from a Tektronix Oscilloscope but tested at 100% (In the CRO it was used as the error amp for a HV regulator using 12B4 pass tubes).

Any advice greatfully received.

Thanks,
Ian

[Enzo]

Elevating is done to prevent the AC heater current from coupling into the cathode. Your DC heaters won't do that, so elevating is not useful, unless you have a CF or something with a high cathode voltage that might exceed the heater-cathode voltage rating.
Someone else may have better ideas, but I would couple the DC to ground with a small cap or directly, but doing both sides of it seems redundant. It makes a different how you are deriving your DC heater current. A bridge at the end of the AC heater string? A separate DC supply?

[Gingertube]

Supply plan - Separate power tranny winding, bridge rectifier + filter cap(s) to a 12V 3 terminal regulator with series diode in ground leg for 12.6 - 12.7 Volts.

Probably use DC heaters for (at least) first 2 gain stages.

Cheers,
Ian

[Gregg]

I would recommend 3).
Floating DC heaters is usually not necessary but since you have two cathode followers in the AFD100 preamp and since many brands of tubes can't handle the CF position very well (some fail instantly) it's a good idea to float the heaters to say 50-80V. The easiest way to do it is to connect the 0V side of the heater DC supply to the 50-80V point.