alternative for heater / B+ references

[cbutterworth]

John Broskie and others strongly urge that the heaters in the Aikido should see the same magnitude of heater-to-cathode between different sides of each triode. The common way of doing this is by referencing the center-tap of the 6.3V trafo to 1/4 B+. The common way of doing this is to use a resistor based voltage divider, with the second resistor paralleled by a 0.1uF cap.

I have done this and consequently, my Aikido's filament trafo sees around 80VDC, which is approx 1/4 of the B+ voltage.

However, when the filaments are being warmed in the absence of B+, the 6.3 trafo is very quiet. When B+ voltage comes on, the 6.3V trafo begins to buzz. Could this method of voltage divider be to blame? Of course, problems may lie elsewhere in my PSU. I have also noted that both secondaries of the 6.3V trafo seem balanced with the center-tap, so I doubt that the center-tap is off-center.

Broskie does mention floating the heater supply via a 0.1uF cap. Exactly how is this done?

NOTE: I am also using DC regulation for my heaters. AND what I really need is an oscilloscope to really investigate my PSU.

Charlie

[Poindexter]

Try a bigger cap. I use 100uF.

Poinz

[rdf]

Sounds like a wiring or filament transformer issue. It might not like DC on the secondary. What value resistors are you using for the B+ split?

The question of capacitor size is an interesting one. Filament to cathode leakage appears to have two mechanisms. Due to the very low coupling capacitance betwen them at low (audio) frequencies it's effectively all emmission and in the general case eliminated by biasing the filament above the cathode. At RF frequencies the capacitance between filament and cathode becomes dominant and DC bias no longer effective, an AC bypass from filament centre tap to ground is required. But, large caps make poor HF bypasses. Lately my preference is to DC bias the filaments roughly 75% of rated cathode/filamant voltage and use smallish film caps between centre tap and ground.

[cbutterworth]

I am using 300K / 2W and 100K / 0.5W with a 0.1uF cap as per Broskie's instructions.

As for DC biasing, I have no idea what you mean - sorry.

Broskie does say that floating the heaters is an alternative. So, how is this done? Do I simply connect an appropriate valued cap between filament center-tap and GND? However, how does this ensure that heater-cathode voltages dont' swing out of whack?

Charlie

[cbutterworth]

ANy ideas if floating the DC filament heaters would work? If it would how do I do it?

Charlie