Isolation

[eeka chu]

This is not strictly with a valve project in mind, but I noticed in Morgan Jone's Valve Amplifiers.

I know that until you reference a floating voltage to a fixed point, like the Earth, the voltage present on that terminal is only virtual, or differential to the other terminal on the same circuit.

What interested me, was his design for a valve heater supply. Firstly, he uses a separate supply for the heaters and the high voltage lines. These both come from separate transformers or windings. So there isn't a differential voltage between them, they're isolated from each other.

Later, he describes the structure of a circuit for referencing the LV supply to the HV supply to reduce noise.

If you were running the heaters on isolated DC, why would you need to elevate their reference voltage?

If they were referenceless in the first place, why would any noise coming from them pass over onto the HV lines? Because there's no referenced differential between them. They're both 'floating' around at virtual points, with only a differential established within their own two polarities.

I can only think that perhaps this might be useful for DHT, where-in the heater and high voltage lines would be tied together. Am I right or am I missing something here?

I understand the words floating, isolated, virtual and what they mean. I spent a long time looking at multiple output SWMPs that used this terminology all the time. I just don't see how it would apply to an IDHT, since the heater and cathode wouldn't have any electrical differential between each other; there is no reason to tie the heater supply to the HV ground, that'd help create problems.

Would this reference for the heaters just be for DHT's?

[audiousername]

Valve heaters should have their DC potential defined. A 'floating' heater supply can cause noise due to heater/cathode leakage. The heater is NOT perfectly insulated from the cathode, it has capacitance and some other things, not to mention that the heater and cathode may sometimes form the two electrodes of a diode. Referencing the heater to some potential and other tricks are used to reduce noise coupling and turn this 'diode' off.

Err.... maybe someone more knowledgeable in this area should comment! This is what I understand anyway.

[cogsncogs]

And don't forget the maximum allowable voltage potential between the filament and cathode in IDHT. Especially in circuits such as cathode followers, SRPP, mu-followers etc. And also there is maximum allowable resistance between filament and cathode. Try to float a filament supply in say a phono preamp or line amp and see how much noise (hum) you get! DC or AC!

Wayne

[eeka chu]

Hmmm, I undertsand what you're saying, I hadn't thought of that!

Thanks!

[Geek]

Originally posted by audiousername
Valve heaters should have their DC potential defined. A 'floating' heater supply can cause noise due to heater/cathode leakage. The heater is NOT perfectly insulated from the cathode, it has capacitance and some other things, not to mention that the heater and cathode may sometimes form the two electrodes of a diode. Referencing the heater to some potential and other tricks are used to reduce noise coupling and turn this 'diode' off.

Err.... maybe someone more knowledgeable in this area should comment! This is what I understand anyway.

Very true.

I mahe a lot of totem-pole circuits and the upper tube has this problem, since the cathode can be 200 to 400V above ground. I eliminate the noise problem effectively by using a floating heater supply for the top tubes and bypass them to ground with a 2.2uF electrolytic of sufficient voltage for preamp tubes, up to 47uF for power tubes.