Voltage doubler heater supply design problem

[trobbins]

Yes, the double doubler type rectifier configuration is only worthwhile when you're 'stuck'. For bias supplies with little power requirement it is just a few small parts so no big deal.

B+ rectifier noise transferring in to a heater winding is somewhat a complex interaction. B+ secondary winding resistance can spread out the current pulse waveform. Primary winding resistance can both also spread out the B+ current pulse waveform, but also increase the primary voltage waveform level of the noise voltages. The heater winding coupling to B+ winding and primary winding, and the B+ winding leakage inductance all play a part.

I recall Merlin showing some hum noise waveform levels that were able to be quite well attenuated using a humdinger pot, compared to a standard preset humdinger.

[Mick Bailey]

I've used the humdinger pot with good success on regular amp builds, even when there's a CT on the heater supply. It's better at nulling the total imbalance in the circuit. I have a Fender in right now where the owner inserted one of the power tubes incorrectly and burnt out the hum pot. I temporarily replaced it with a fixed-value ground reference but it wasn't as hum-free as with the pot (which was set considerably off-centre). In the case of the doubler I couldn't get a reference point for full-wave, and the 1/2 wave is already tied to ground.

[jpfamps]

Whenever I've built similar devices I've always used a 12 VDC supply and a discrete charge pump for the HT.

This works very well, with no hum from the filaments.