Lead dress for tubes with internal shields

[cdemike]

Wondering if there is any opinion one way or another on how to make the most of internal shields in tubes like EF86, 5879, EBC81, etc. The case I'm most interested in is whether it makes a difference if I solder a cathode resistor and bypass capacitor directly from the cathode pin to the shield pin at the socket. That would let there be only a ground wire coming from the socket rather than a cathode and a ground wire, so my thinking is that, on the one hand, it could reduce noise by reducing the total length of leads running to and from the socket. On the other hand, that set up might hurt more than it helps if there is enough interaction between the cathode and the shield, especially if the cathode is bypassed or the lead from the shield pin to chassis ground is particularly long.

I haven't encountered any issues in the amps I've tried this in, but I haven't had a chance to compare the two setups under a wide variety of conditions. Under conditions with a lot of EM interference (I'm thinking nearby like nearby neon signs, wifi routers powerful enough for commercial use, and other sources at particularly noisy gig venues), does one set up work better than the other?

[R.G.]

Hmmm. I hadn’t thought about this. My first knee-jerk is that it depends on what the tube designer put that shield in there to shield against. For instance, was it to shield against internal capacitance and plate swings, extend RF response, or to shield out external RF. You’re pretty much stuck with shot noise and thermal noise, so it can’t be that. I’d have to research the tubes in question for a more complete picture of it.

My best guess before l go look it up is that in audio applications it won’t matter. In that case, tying the internal shield to the local cathode ground is probably fine, as the shield should not be carrying much current; and resolving the shield’s current in a small local loop is probably better in terms of not picking up other stuff magnetically or capacitively.

I have a copy of the MIT Radiation Labs book series for electron tubes. I’ll give a read and see what I can find on internal shields.

[pdf64]

I suggest to connect it to the nearest convenient point of the chassis metalwork, eg a solder lug at its valve socket fastener.
By that, any RFI nasties it picks up will be kept as far separate from the amp circuit as is feasible.