What your saying here seems contradictory .If as you say the input's to IC3 A (the cathode of D15) are not changing when you select ac or dc mode (u measure a constant 56 mV ) .But the output of this opamp (pin 1) still adjust in ac mode correctly . (and so it would in that event function normally anyway ).
Then when you switch into dc mode (still 52mV at that D15 cathode node)then pin 1 will no longer adjust using balance cntrl (which is also what we would expect for normal function in dc.). So in effect your saying this cntrl's output voltages seems to be working fine but the input's to it are broken . Thats a bit contradictory really
What also makes it odd is that given the schematic you posted that output behaviour doesn't seem to be possible with that D15 node stuck at 56mV .When balance is adjusted to give output (pin 1) -15V in ac mode what is the V at pin 2,3 and the balance cntrl output pin ? What Does the bottom of this balance cntrl connect to ?(I was thinking the small sloping line symbol here refered to a local case gnd ?) .
Check OP IC3B pins 6,5,7 ,given what it is , ic3B is meant to be either full on or full off ,it goes back to M1 connector what goes into this conector ? is this supposed to feed a simple dc level then a continuous square wave in ac mode ?. Is there a part of the schematics missing with your dc mode switching..etc on it ?
Given the output's from this cntrl seem to be in normal spec what would we hoping to achieve ? except to solve a puzzle .
Regards