Originally Posted by
Barry
Yes, the interference can only be caused by near adjacent cables, but to have an effect the screening of the audio cable would have to be poor. Also since in a mains cable the L and N wires are adjacent and run very close together within the outer sheath, separated by their own insulation, the fields they generate are almost mutually cancelling.
Unless there is a fault with the mains wiring feeding the wall sockets (possibly due to some sloppy DIY work), again the field generated by the cables will be virtually non-existent. In the UK we still use the "ring" circuits for the mains wiring. In Europe and elsewhere, "radial" circuits, or "tree and branch" circuits are used, and these force the return current to travel back down the same piece of cable, hence the fields will be self-cancelling. In the UK's ring system, the wires are laid out in a circle with both the Live and Neutral wires starting and finishing at the consumer unit. This means that the current used from a socket outlet has two possible ways to flow to and from the consumer unit. Currents many not flow equally both ways around the ring, so the magnetic fields produced may not cancel, and the cables then radiate these higher magnetic fields into the room and the adjacent rooms.
At every socket on the ring two cables are joined. A small difference in the impedance back each way to the consumer unit around the loop will cause a problem almost as bad as a poor connection. In the worse case, which is surprisingly common, a Line or Neutral actually gets disconnected (often as a result of incompetant DIY) and this effectively results in a single turn transformer loop. When a significant load is applied, the magnetic field in the rooms inside the ring can rise to many microTesla. To put this into context in the UK the average for the mains cable feeder into the house is 0.04 microTesla. For detached houses set back from the road it is about 0.02uT. Flats and terraced houses right up to the pavement have higher fields, however if they are about 0.1uT then an external supply fault should be suspected.
Consider a ring running around a 3 x 3 metre room supporting a total load of 2A (i.e. 450W). The radiated magnetic field 60cm into the room will be 0.01uT, falling to 0.001uT one metre in. Whereas if both L and N wires were to become disconnected (worse case scenario) the respective magnetic fields would be 0.6uT and 1.5uT.
Another source of unwanted magnetic fields (at frequencies higher than 50Hz) are those generated by 'wall warts'. Commonly used to charge mobile phones (but sometimes audio items), these device can emit high fields nearby, and also inject noise back into the mains system. If not in use they should unplugged, or at least switched off at the socket.
