Strange Bassman 50 Issue

[boyer]

Hey gang,

Just kind of polling to see if anyone's had a similar issue to the silly one I'm having right now.

I've got a 74 Bassman 50, which had been blackfaced when I bought it. I've done several modifications to it, including:

* extra gain state in first channel
* Mid pot (both channels)
* More guitar appropriate coupling caps
* Channel switching via foot switch or the re-purposed Deep switch (using a DPDT relay. The centre pole of one switch is connected to the node feeding the PI, and a coupling cap from each channel feed is conected to either side.
The other half of the switch grounds out which ever channel is not being used at the coupling cap node)
* first channel MV (master volume is a 1m log between the coupling cap to the PI feed and the relay)
* 150R/15W across speaker output, just incase


So, having done all that, I've spent a lot of time inside that amp. Here's my trouble:
Some how, the first channel is making it's way into the PI, even if it's grounded out, or even disconnected at the third stage!
I've completely lifted the coupling cap from the 3rd stage to the PI, and there is still 1st channel signal there.

When the amp is switched to the first channel, the second is grounded out. Everything seems to work well. When switched to the second channel, the first is grounded out. The second channel works as expected, except that the volume controls of the first channel have a dramatic effect on the volume! The tone controls of the first channel do not effect the second channel.... Since the tone controls don't effect things, I'm thinking it's gotta be coming from one of the first two stages. (ie before the tone stack and recovery stage)

I think there MUST be an under board wire, or something like that. There is NOTHING I (and two other people, just to make sure I wasn't over looking) have been able to see in months of looking.
My father seems to think that it's inductive coupling somewhere along the line, but I don't think I buy that... All important signal cables are as short as possible, shielded with braided shielding, grounded at one end nd are as far away from any trafo's as possible. Gotta be something physical.

Of course I have the one bassman which there isn't a layout floating around for! I've lifted up the board as much as possible, but can't see anything under there.

I'm down to the point now where I'm very seriously considering taking the entire first channel off the board and checking for shorts between eyelets. I dunno what else to do....

[dai h.]

if you've ruled out signal bleeding in through wiring and have everything isolated that should be (with the use of the shielded wire) coupling can happen through sharing of stages using the same dual triode. Have you considered that possibility?

[Enzo]

First thing to my mind is insufficient decoupling in the B+ supply. Perhaps give each channel its own node or nodes.

When yu say a channel is grounded out, what does that mean? You ground the output of the channel? How about grounding the input of the channel That wioll prevent any signal being amplified in that channel when not selected. Nothng amplified means nothing can leak out. If you merely ground the output of a channel the tubes and things earlier in the channel are still working hard as ever.

[boyer]

Thanks for the replies guys.

The two channels do not share any tubes. This was one of my first thoughts actually.

When I say each channel is grounded, yes, I'm referring to the output (the node where the mix resistors at the PI used to live).
I am only using one input jack, and running that to the grid of the first stage. I run a shielded wire from that grid to the grid of the 4th stage (first of the 2nd channel).

I'll have to come up with a different input/switching scheme to accommodate grounding the input of the stage. Any recommendations?

Perhaps I should take the feed from the 2nd channel from the plate of the first stage, through a coupling cap and dropping resistor to an empty node, and ground that node when the channel isn't in use?

Also, you might be onto something with the B+ coupling. I'll scope it tonight.

Thanks again guys!

[boyer]

RE: Channel Switching

Hmm, thinking about it more, I'm thinking for the channel switching, I should make a node off the plate of the first stage, and feed the relay from there.

Where exactly should I ground though? Is it okay (or preferable at least) to ground the grid of the 2nd stage of which ever channel is not in use?