There is a coil, and a magnetic field passing through it. A sheet of superconducting material is inside it, covering most of the inside of the coil. (E.g. coil and sheet same plane). When the sheet is in superconductive state, it blocks most of the magnetic field due to Meissner effect. When it's not, it doesn't. So if we can switch the sheet between super-conductive and non-superconductive states we can generate electricity in the coil. If we place a second coil, perpendicular to the sheet, and going around it, we can use it's magnetic field to switch between the two states. When the magnetic field of the second coild exceeds the critical field, it will force the sheet in non-superconductive state. However, the sheet doesn't cover the hole of the second coil, it only splits it in half. Therefore the sheet will only block a very small fraction of the magnetic field coming through it. So we get less feedback voltage in the second (primary) coil than we produce in the primary coil. It is multiple orders of magnitude difference if the sheet is very thin. The two coils don't directly produce voltage in each-other as they are perpendicular with common center. So what we essentially have here (theoretically) is a transformer that doesn't conserve energy. Please note I haven't experimentally tested this device.
http://rogercortesi.com/eqn/tempimagedir/eqn3018.png This is the energy creation formula, derived directly from the vacuum version of Poynting theorem by substituting the total current with the sum of free current and bound current. Here U is the total energy in a closed system. This formula can be used to prove the above device does produce energy.