Ok, if you still have the sound, then I still think you have contact between the stationary field and the rotating armature. That is just the sound I would expect with that. You should not have that sound. The motor should be virtually silent. When you check for contact you have to remember that the armature is going to try its hardest to contact the stationary field in some manner. The motor works because magnetic fields are fighting with each other: things are not going to stay in alignment unless constrained. When you check for contact, try to push the armature as far our of alignment as possible, rotating with your fingers, and see if you can't find a contact point. It need not be a large area, perhaps just a burr somewhere, or a corner. Looking closely at this photo, I see a suspicious shine on the top corner of the armature segment. Looking even closer, I see faint abrasions, not yet through the paint, on its leading edge.
@Junior posted:
It may be that reducing the vertical play is enough to fix this, since that will also reduce the lateral deviation slightly (imagine the tilted armature shaft as the hypotenuse of a triangle). If it still persists, then I would be looking for an oversized bearing, or else out-of-alignment plates in the stationary field. Some file work might fix the latter problem. Or a hammer (carefully, of course!).
