I would check around that D5 protection diode. The 1N400x may be ok, but it's not unknown for the PCB track to have fused out instead. A parallel protection scheme cannot be proved to be present and effective unless you test for the reverse path via the diode.