Cathodyne PI help -- no plate/grid/cathode = sound?

[Paultergeist]

All,

Just trying to wrap my brain around this: trouble-shooting an old amp....it has semi-intermittent noise which seems to be isolated to the area of the circuit where the PI lies. Pull the PI (driver) tube = no more noise but no guitar sound either.

Anyway, I am trying to isolate the noise, so I have been lifting and disconnecting wires in an effort to figure things out. Bottom line is this: if I disconnect the plate, grid, cathode from the PI side of the tube (other side is a gain stage), there is STILL guitar signal being conveyed through the amplifier. The signal is pretty weak (also totally clean) -- with the volume pot dimed, the loudness might be on the order of 3 x acoustic steel-string guitar -- but the amp plays.

Does this make sense? Lift every lead at the PI, and the amp still plays? Or, is there some sort of bleed-over of signal which may explain the under-lying noise problem?

Thanks for any help.

[tubeswell]

If a filter cap node goes, it can cause the HT to 'become' the signal path - as a result of not having the supply resistors at that node properly decoupled at AC (which could account for the faint signal sound with the PI plate etc disconnected). It could also account for 'ticking' (as a sort of oscillation between stages that is facilitated by un-decoupled supply resistors). Just guessing' (as I'm only going by your brief description on the internet and I'm miles away from you and therefore can't be on the spot to assess the situation myself ) Of course I could be completely guessing up the wrong tree.

[rdjones]

If there's NFB, signal could be following the path from the pre-PI driver to the OT secondary.
It'd be pretty low though.

rd

[Paultergeist]

Thank you for the replies.

At this point in time, I think -- emphasize "think" -- that I may have found the actual problem with the amp -- not the PI after all.

That said, I am thinking that it must be more or less normal -- at least in some circuits -- for there to be some low-level of coupling through the ground path or power supply rail for signal to get to the power tubes -- even with the PI disconnected.

Thank you both for weighing-in on this.

[tubeswell]

If the power supply dropping resistors are sufficiently decoupled, then the HT should be steady and 'seem' to AC to be one and the same as ground.

But if the droppers aren't decoupled, then the signal swing at the plates will be using the (plate resistors and the) dropping resistors (combined), in which case, the signal swing across the dropping resistors will appear in other stages