[dekhn]

Well, you still lose current over time... for example, we had to dump a bucket of electrons into our superconducting, supercooled magnet about every month ago to keep things swirling properly.

(The EE I worked with later didn't believe me. See https://en.wikipedia.org/wiki/Superconducting_magnet#Persist... and note that the loss was due to details of magnetic superconductors, not superconductors in general)

[Sharlin]

Yeah,

> Experiments have demonstrated that currents in superconducting coils can persist for years without any measurable degradation. Experimental evidence points to a lifetime of at least 100,000 years. Theoretical estimates for the lifetime of a persistent current can exceed the estimated lifetime of the universe, depending on the wire geometry and the temperature. In practice, currents injected in superconducting coils have persisted for more than 27 years (as of August 2022) in superconducting gravimeters.

(from https://en.wikipedia.org/wiki/Superconductivity)

[klyrs]

If you can extract work from the field generated by the supercurrent, it must come from somewhere. Small supercurrents make small fields, so things like adiabatic CPUs seem interesting.

[OkayPhysicist]

Can you extract work from a constant magnetic field? As I remember my physics education, constant magnetic fields don't do any work, since they apply a force perpendicular to the direction of motion.

[klyrs]

I'm not a physicist, but I play with circuitry... two nearby loops of wire are magnetically coupled. If one is a superconductor with some stored current, and the other is a normal conductor with some resistance, then it stands to reason that the supercurrent will burn heat off in the resistive loop.

[OkayPhysicist]

What happens in the scenario where the superconductor coil is just replaced with a permanent magnet? I'm pretty sure the energy that the loop dissipates comes from the energy required to move the loop into position, which the inductor experiences as a changing magnetic field.