Looks as though the "monostable " timing is from R254 and C230, set by
the output swinging fully negative and reset after the cap discharges.
Normal bias Q213(B) is -1V and goes to -3.3V in DC swing state.
Don't know what is causing the initial disturbance to the bias, but
previously with amps used or stored in damp conditions, I've found it
due to leakage in minor Qs. Owner said its stored indoors and never used
outdoors but rust spots on metalwork and some pcb CuCarbonate green
spots, say otherwise.
Looks like some suck it and see zener clamps to localise down.