Minimizing noise on series/stacked filaments

[oxbow_lake]

Hello!

For fun let's say I wanted to build an non-reverb Princeton-like amp using 6N2P and 50L6 tubes, and let's say I have a 115-60V transformer that I'm going to use to supply the heaters. In theory I could stack the filaments like this:

62167448-e630-409a-ba70-bc06ba17fd20~1.jpg

Questions:
1. If elevating the heater voltage through an artificial center tap, how high do I want to elevate it? 60V? Can I use a humdinger pot just like if they were normal 6.3V filaments?
2. Is this the lowest-noise stacking arrangement, with the 50L6s on one end and the 6N2P with the input stage closest to the other end? Or is it better to sandwich the 50L6s between the 6N2Ps?
3. Is this just stupid? I mean, yes, it's just stupid, otherwise there would be tons of 50L6 push pull amps out there which there aren't. But I'm interested in hearing smarter people tell me why!

Thanks!

[martin manning]

I would try putting the faux CT with 2x 250R or 500R hum dinger pot across the input 6n2p to center its heaters around zero volts. The 50L6's will be elevated about 35V. The heater currents are not very well matched, so the 6n2p's might be running at ~5.3V instead of 6.3V.

[oxbow_lake]

Thanks!

Since the filaments are not simple resistors I'm not really sure how you estimate what the final current/voltage will be when they're in series - how does that work? I had naively hoped that the total current flow would end up close to 300mA, which is pretty close to within spec for the 6N2Ps.

[martin manning]

Yes, the resistance varies, but I assumed they are fixed. Good enough for this purpose, and regardless of the voltage the input tube heater runs at, the faux CT will center it around zero.

[oxbow_lake]

Cool. Is there any benefit tying the CT to a low positive voltage, i.e. the cathode resistor if I cathode bias the output tubes?

[martin manning]

I don't think so, but you could try it and see. Before doing anything else I would run a bench test to see if the heaters will run within 10% of spec voltages.

[Astronomicum]

Those 50L6s need to be wired in series, not parallel. Not sure how you would propose wiring the humdinger but while 500 ohms works for 6.3V, you would smoke 500ohms at 115V. Showing the full arrangement here is important.

[sluckey]

Those 50L6s need to be wired in series, not parallel.

Not in this case. He proposes to use a 60VAC transformer.