safety with high voltage input capacitor ?

[Bigun]

I have a circuit in mind which uses a negative B rail instead of a positive B rail. This would put the cathode of the input tube (common cathode) at B- and therefore the grid would also be close to B-. The input signal would be decoupled from the grid by a high voltage capacitor. Only this capacitor would prevent B- from appearing on the input RCA connector. Is this a big no-no ?

The alternative is an input transformer - is this any better ?

 

[kruesi]

Low leakage film and foil type with extended foil may be best here... Also a very reliable pulldown resistor directly across the input- just a suggestion, likely obvious. You should be fine otherwise. Here's the place to spend a few bucks on high quality components.

 

[funk1980]

Coupling caps are there to block DC. It doesn't matter if it's negative or positive. It's all about reference, meaning it makes no different for the cap to go from 100V to ground, than it is to go from ground to -100V, since it'll be the same potential.

 

[M Gregg]

Where are you going to reference the input to? Earth?

Depending on the value of the cap it will charge up how does it discharge after power off?

Polypropylene comes to mind with a very high WV. Its not something I would do.
An input transformer gives isolation as long as the insulation rating is high enough. Primary input and ground/earth, secondary output to circuit and no discharge to think about. However they are expensive.

Why can't you drive the input from a standard tube section if its a cap input?
Its only like having a preamp stage in the front..🙂<<its only the same as the bottom half of a futterman OTL

Regards
M. Gregg

 

[funk1980]

Since it's a coupling cap to a common cathode stage, I assume it'll be of a very small value. So (dis)charge shouldn't be an issue.

 

[M Gregg]

Since it's a coupling cap to a common cathode stage, I assume it'll be of a very small value. So (dis)charge shouldn't be an issue.

What if its a self healing capacitor? If not enough current flows to clear the fault..(just a thought)

Regards
M. Gregg

 

[Conrad Hoffman]

Design for cap failure. I'd want some series resistance to limit current, and some protection diodes to protect whatever is connected to the input from both cap failure voltage and spikes on power-up/down. Also, a bleed resistor to keep the input at ground potential. Then I'd want a fuse or something in the HV supply line to protect the supply from the protection circuit. You really want to think this one through!

 

[Bigun]

The reason why I'm considering this type of input is because I am exploring some hybrid circuit options in which the output from the plate should be at dc ground potential.

The challenge of this type of circuit is the PSRR of the B- supply, but that's just engineering.

The attached is what I have in mind, and it includes what I think are all the suggestions you guys have made so far. There is a series resistor with the input cap to limit current in case of a cap-short-failure, there is a resistor to ground at the RCA input connector to shunt current to ground, and a pair of zener diodes to clamp the input to within a fixed voltage from ground - say +/- 12V. I would always fuse my supply rails too.

 

[Conrad Hoffman]

Seems reasonable, but I'd put the zeners in series, as otherwise they'll break down like regular diodes, clamping the input to a low value.

 

[Bigun]

Seems reasonable, but I'd put the zeners in series, as otherwise they'll break down like regular diodes, clamping the input to a low value.

You're right! that was a big oops on my part, they need to be in series 😱

 

[Mescalero]

If shorting is a concern, try using "Y" caps. They are high quality film caps and are rated in AC volts. To determine max working DC volts, it is customary double the AC rated value.