I was discussing high-quality twisted pair cables that go from a DSL modem to a wallsocket and have an overall shield.
A friend has pointed out that shielded cables ought to have the shield connected to earth. My understanding of this point is that it prevents the local potential inside the Faraday cage possibly bouncing up and down relative to the outside environment. Is this an accurate interpretation? I am wondering how much benefit you get anyway in an un-grounded setup, in this case where the conductors inside are a twisted pair. Since the inside of the shield is a faraday cage the potential is flat and the electric field is zero i.e. grad V = E = 0 everywhere so this surely makes life say for the twisted pair. Maybe the charge redistribution response time of the shield is inadequate or it has internal resistance so that the interior is not an effective Faraday cage up to sufficiently high frequencies. (Superconducting shields are out of my price bracket.)