DC heaters question

[Helmholtz]

Measure at the speaker output.
If there's no measurable ripple, you don't have a problem.

[catalin gramada]

Here is the speaker output with all controls dimmed. There is still no sign of 100hz ripple.

[Helmholtz]

So, what's your problem?

[catalin gramada]

There are no evident problems in both instances. That mean both variants of grounding are correct and no one take advantage over the other ? So, there is a reason to choose one against the other please ? Other than two resistors savings ?


Late : if I inject more ripple from the source I may do a comparative test ? How big should be the ripple level to be evident in a grounding test to see a difference please ? Thx.

[catalin gramada]

And think I may answering myself. The minimum noise I can get with a mix of dc and substantial 1v ac ripple was with the negative tied to ground. But again with 2 mv AC will not show any diference...that doesn't mean it don't.

[Helmholtz]

2mv peak ripple equates to an ablolute maximum 2mv ripple on the anode, assuming the heaters are short circuit to the cathode, which is unlikely.

2mVp between cathode and ground would be amplified to around 100mVp at the anode using the tube in common grid mode.
But as you said, a cathode-to-filament short is extremely unlikely.

[catalin gramada]

Hello.
Because my winding have just 0.2 ohm resistance I tried a diversion to limit the charging current and I instaled one ohm resistor between bridge and first capacitor. To get same voltage output for same current as in first case (when I did not used 1 ohm resistor but only 7.5 ohm between caps) I used a 5 ohm resistor. The ripple goes to 2.5mV vs 1.5mV in first case....(dmm measured)
I expected to use 1 ohm + 6.5 ohm resistors and almost same ripple for same current, not 1+5 ohms and almost doubled ripple...I don't get it...is not linear...? It is normal or I have some measurement errors somewhere and have to check more carefully please ?

[Helmholtz]

I guess the reason is different ripple currents in the separate resistors, so it's more complicated than just adding resistor values.
I would use PSUD2 for simulation.

[catalin gramada]

Yes, right, first one is not part of dc operation, thanks. One more please: the capacitor max ripple current is specified as peak or rms current please ? Thx.

[Helmholtz]

The ripple current limit is specified as RMS value.
The limit corresponds to a max. internal temperature rise due to Ploss = Irmsē x ESR.