[MrFixr55]

Sorry for the wordiness.  My humble $0.02:
I never understood why just about every AA5 radio has the power switch between "B-" and the power cord. Is this to reduce hum?
In radios of this vintage, before metal tubes, the chassis was usually connected to B- via a cap or cap and resistor to get the chassis at B- potential for RF / AF but not for power.  This is to reduce the possibility of electrocution before the advent of "double insulation", interlocks, etc.  In act, on the latest AA5s I believe that the chassis is connected to B- because of double isulation and interlocks.  Have to check the schematic for a late model.  (You experts know all of this; I make this comment for Pholks new to the hobby.)

Prior to metal tubes, the IF and 1st AF tubes had metal can style shields that connected electrically to chassis GND.  If the 1st AF tube had a grid cap (75, 6Q7, etc.) the grid wire was shielded or was also inside the can shield.  Otherwise, the 1st AF stage was very unstable.  Even in AA5s with miniature tubes (12AV6, etc.) that tube requires a shield connected to chassis.

Verify the following:

• Ensure that B- is at "Neutral" level and not "Hot" level.  Some restorers use a polarized plug and place the power switch in the hot lead. The issue with the switch being in the B- leg instead of the "hot leg is that if the power plug is inserted so that switch is at neutral, with the switch open, the chassis is "hot" through the filament string and chassis is at neutral only if the switch is closed.  Visa versa if the power plug is inserted so the switch is in the "hot lead of the power line.  With the switch open, the chassis is cold but with the switch closed, the chassis is hot.
  
• For this post, B- is the negative of the B power supply terminal and the "low point" of the series filament string.  GND is chassis ground.  In most AA5 sets, there is either a 
  
• Verify that the cap that connects B- to GND is the correct value.  In the original schematic, this is the 0.05uF cap between the "black" lead (B-) and the cap can (GND) of cap a**'y #7.
  
• Ohm between the metal envelope of the 12SQ7 and its pin 1.
  
• With the tube in place, ohm between the tube envelope and chassis.  This should be 0 Ohm.
  
• Verify continuity between GND, Pin 1 of the 12SQ7, 12SA7 and 12A8 (assuming that you moved 12SQ7 Pin 1 back to Chassis GND).
  
• Verify B- connections at 12SQ7 socket for filament, cathode, grid resistor return, etc.  In this mod, did you retain the circuit as drawn for the 75 in the original schematic?  I don't think that the 12SQ7 requires negative bias but if the grid resistor bias circuit was retained, ensure that the resistor is bypassed by the correct cap.  If this was replaced by a cathode resistor, ensure that it is bypassed by a cap. (In RCA sets using a 12SQ7, the cathode and 4.7 MOhm grid resistor are directly connected to B-.  If you retained the 75 circuit, this may be the issue.)
  
• If only 1 diode of the 12SQ7 is used, connect the other diode to the first diode or to B-.
  
• If all else fails, swap out the 12SQ7.
  

Hope this helps.  I have gotten my "U know what" kicked by simple issues on simple radios on more than one occasion!