Signal coming through tube amp with volume down

ok....the problem seems to center in between gain stage 3 and 4.  If I short to ground the grid of gain stage 4 (post 1M resistor) then all bleed is gone.  If I short pre 1M resistor then the problem is mostly gone. 
Ideas?

audiomixer said:

is that because of rather high signal impedances rather then because of level that you get bleeding there?

- michael

Click to expand...

High impedance is the main factor. If the signal is low level there tends to be more gain after it and vice versa.

Cheers

Ian

figuringstuffout said:

ok....the problem seems to center in between gain stage 3 and 4.  If I short to ground the grid of gain stage 4 (post 1M resistor) then all bleed is gone.  If I short pre 1M resistor then the problem is mostly gone. 
Ideas?

Click to expand...

I had not noticed all those large value series resistors before. What are they for? With Miller effect they will seriously cut the HF end of the response. I know guitar amps have them often at the input but are thy really necessary in every stage?

Cheers

Ian

ruffrecords said:

figuringstuffout said:

ok....the problem seems to center in between gain stage 3 and 4.  If I short to ground the grid of gain stage 4 (post 1M resistor) then all bleed is gone.  If I short pre 1M resistor then the problem is mostly gone. 
Ideas?

Click to expand...

I had not noticed all those large value series resistors before. What are they for? With Miller effect they will seriously cut the HF end of the response. I know guitar amps have them often at the input but are thy really necessary in every stage?

Cheers

Ian

Click to expand...

With as much gain as this amp is using it is actually important for a couple of reasons (doesn't have to be quite 1M, but higher than 220K in my experience).  1.  The high frequency rolloff is important in order to keep the ultra high end from becoming uncontrollable  2.  When I take out the high level resistor out the distortion becomes very flabby and nasty...kind of like when a tube is overdriven to the point that it sounds like it is falling apart.

I just want to know why so much signal bleeds through this circuit even when the master volume is all the way down unless I short to ground this junction, post 1M resistor.

It is the nature of high impedance designs that the high impedance nodes are susceptible to signal coupling in.

Physical layout is important.

JR

ok, further update.

Yes I am dumb for not having thought this was important, but I actually have a 220K resistor in series after the Master Volume pot right before the input into the PI.  I had left it there after messing with two different channels a while back (Resistive Mixer).  At any rate, I removed the 220K resistor and the problem mostly went away.  So I hooked up a 1M pot as a variable resistor to see what was going on (in series after the Master Volume pot) and low and behold, as I increase the resistance between the Master Volume output and the 22n cap going into the PI the signal bleed gets worse until it is actually quite loud. 
Any ideas as to why this might be happening?  When I ground the input of the PI before the 22n cap the problem mostly goes away and when I remove the series resistor the problem mostly goes away.
Oh, yeah, and when I completely unplug the input going into the PI and leave the 22n input cap unconnected to the Master Volume cap, the exact same signal bleeding  problem happens again.

With the volume control turned all the way down, what's the resistance between the wiper lug and signal ground?

The resistance is zero ohms. When I disconnect the PI input from the master volume pot the resistance at the 22n cap going into the PI is infinity ohms