[MrFixr55]

Sorry, I know that this is 4 months late but here goes:
The 82 schematic is interesting to say the least. Look into the 24A that is referred as the CONT (Control?) tube, This may be an AVC circuit. The volume control is part of a resistor "Ladder" Voltage divider for various B voltage, Check for voltage variation at either end leg of the volume control and all along the "ladder" for changes in voltage. Replace questionable resistors and caps. Check for changes at the wiper. However, the wiper voltage may change as the "Control" Tube warms up. Read on.

The volume control varies the cathode bias of the "Control" tube, which varies grid bias on the 1st detector (Mixer) and the IF Amp. Based on where on the voltage divider the cathodes of the IF and Mixer and the plate of the Control Tube are, the plate of the Control tube is negative with respect to the cathode. The more current that passes through the tube the more negative the plate gets in relationship to the Mixer and IF Cathodes, making the grids of these tubes more negative. Cathode voltage of the 2nd detector controls the grid voltage of the Control Tube, therefore AVC is obtained. As the current of the tube goes up, the bias on the mixer and IF amp is increased. Could be that slow warmup of that tube is the issue. Swap with the 2nd detector, or dig up a 24A from somewhere else. The 35 is a somewhat similar tube but is a remote cutoff (Super Control) tetrode, as opposed to the 24A which is the standard sharp cutoff tetrode tube.