[R-390] filters and wandering gain

[quartz55]

I think I'm getting somewhere with the wandering IF gain.  Injected 455 into grid of V504 and then V503.  No wandering gain.  When I injected it into V502 got the wandering gain.  OK, set IF gain at max so I didn't confuse myself with the R519 gain control.  Changed C517.  No difference, still wandering gain.  Moved injection to input of IF module, removed filter cover.  Notice that as the gain changes, the tuning of the filters change with the variable caps.  Also when I touch the adjustable cap on the 4K filter, I always get a 'snap'.  Took a good look at the soldering around that area, and re-soldered the outputs of the 4K filter with a good hot iron.  Also found a bad solder joint on the inside of the variable 4K cap.  Soldered that better.  Now I seem to have no wandering gain, but I do get a large snap when I touch the 4K filter output connection with the tuning tool which only has a tiny metal piece at the end of a long plastic handle.  I did get the gain dropping and when I touched the 4K filter output, it 'snapped' and gain went back up. It's holding for now.  Wondering if the filters are changing or sometimes not letting the grid of V502 return to ground somehow?  Still everytime I touch the 4K output connection with the tuning tip, I get a 'snap' and the noise level goes way up.  None of the other filters do this.

Is there a direct connection from the input to the output on the filter connections?  Exactly how is the grid of V502 returned through the filters and R507 to the AGC line, just the filter output winding?  If connections are getting loose inside I could see the grid voltage on V502 getting strange. and changing the tube gain.

Ideas here?  I'll have to let it cool off tonight and see if the gain still wanders around from cold, it's been on for a few hours now.

N3DT