5 ms·

A magnetic levitation demo indicates a possible Meissner effect (which is refuted here by this paper), and this is so much easier than a resistance measurement,

A magnetic levitation demo indicates a possible Meissner effect (which is refuted here by this paper), and this is so much easier than a resistance measurement, which is a delicate task and susceptible to experimental errors. So levitation is used as a quick "first check", especially when you don't have a full lab setup to do other tests, which are much more complicated (in addition to a levitation test or a resistance measurement, one can also measure its frequency-dependent magnetic susceptibility curve, temperature-dependent specific heat curve, or its Josephson effect). One cannot measure superconductivity with an ordinary ohmmeter. Electrodes, wires and the ohmmeter itself are resistive, the meter can never show zero ohms. You need to set an experiment up with a current source and a voltmeter to measure the IV curve across the material [1]. Even then, the voltmeter will never show "zero" volt because of noise, which needs to be minimized in the experiment and removed during post-processing. In this case, sample preparation is also problematic as the synthesized material is not uniform. And your measurement is also expected to be temperature-dependent to show the material's transition to a superconductor. It's more involved than seeing it floating on a magnet, to say the least. [1] It's basically the same 4-wire Kelvin sensing used by all milli-ohmmeters. But to characterize superconductivity, you need to do even better. https://en.wikipedia.org/wiki/Four-terminal_sensing https://en.wikipedia.org/wiki/Four-terminal_sensing

haldujai 3y ago
Do you have any idea on what the ballpark cost is to do these replication experiments I keep reading about? Or the more complicated experiments you mention? Asking out of curiosity as I would have assumed it's very expensive but there seem to be many groups of varying sizes doing these and quite quickly which I would have thought would be extremely wasteful but may suggest it's more economically feasible than I had thought.
p-e-w 3y ago
> One cannot measure superconductivity with an ordinary ohmmeter. But one can measure electrical resistivity, and if the material's resistivity is much lower than that of ordinary materials, that would be super interesting and valuable, regardless of whether it is a "true" superconductor or not. If this relatively cheap material had a resistivity around 1/100th that of copper, that would be an engineering breakthrough, and determining that doesn't seem to require a super complicated lab setup. Do we have an upper bound on LK-99's resistivity, and is that upper bound substantially lower than the resistivity of typical conductors?
- chmod600 3y ago
  I think you are missing the point: a typical ohmmeter would have metal probes, so with a superconductor you would just be measuring the resistance of those probes. Any attempt to overcome that problem probably introduces a lot of chances for error or noise, or require some huge sample of the superconductor.
  - spiderfarmer 3y ago
    Zero resistance would be hard to proof this way, but why can't you compare measurements of LK99 with known conductors?
    - foven 3y ago
      You absolutely can, and as you can see in this paper it is absolutely do-able to measure the resistance of these materials. They measure an insulator with something on the order of 10^6 ohm m. For reference, copper has a resistivity of ~ 10^(-8) ohm m, so there are about 14 orders of magnitude of difference. This does make the contact resistance pretty negligible in this case too.
    - 1970-01-01 3y ago
      I have the same thoughts. I think it's because they are not EEs and want to play with the material. A sensing resistor is trivial. https://www.digikey.com/en/articles/fundamentals-of-current-measurement-part-1-current-sense-resistors https://www.digikey.com/en/articles/fundamentals-of-current-...
- nullc 3y ago
  if you were talking about a meter long segment of pure LK99 you'd have a point. But the pieces in question are so small that even if the LK99 were just copper their resistance would be so low that experimental error in a casual test would be substantial compared to the resistance. Measuring the resistance would also be frustrated by the superconductive material being scattered in the mass separated by highly resistive impurities, making it easy to get a false negative.

haldujai 3y ago

Do you have any idea on what the ballpark cost is to do these replication experiments I keep reading about? Or the more complicated experiments you mention? Asking out of curiosity as I would have assumed it's very expensive but there seem to be many groups of varying sizes doing these and quite quickly which I would have thought would be extremely wasteful but may suggest it's more economically feasible than I had thought.

p-e-w 3y ago

> One cannot measure superconductivity with an ordinary ohmmeter. But one can measure electrical resistivity, and if the material's resistivity is much lower than that of ordinary materials, that would be super interesting and valuable, regardless of whether it is a "true" superconductor or not. If this relatively cheap material had a resistivity around 1/100th that of copper, that would be an engineering breakthrough, and determining that doesn't seem to require a super complicated lab setup. Do we have an upper bound on LK-99's resistivity, and is that upper bound substantially lower than the resistivity of typical conductors?

chmod600 3y ago
I think you are missing the point: a typical ohmmeter would have metal probes, so with a superconductor you would just be measuring the resistance of those probes. Any attempt to overcome that problem probably introduces a lot of chances for error or noise, or require some huge sample of the superconductor.
- spiderfarmer 3y ago
  Zero resistance would be hard to proof this way, but why can't you compare measurements of LK99 with known conductors?
  - foven 3y ago
    You absolutely can, and as you can see in this paper it is absolutely do-able to measure the resistance of these materials. They measure an insulator with something on the order of 10^6 ohm m. For reference, copper has a resistivity of ~ 10^(-8) ohm m, so there are about 14 orders of magnitude of difference. This does make the contact resistance pretty negligible in this case too.
  - 1970-01-01 3y ago
    I have the same thoughts. I think it's because they are not EEs and want to play with the material. A sensing resistor is trivial. https://www.digikey.com/en/articles/fundamentals-of-current-measurement-part-1-current-sense-resistors https://www.digikey.com/en/articles/fundamentals-of-current-...
nullc 3y ago
if you were talking about a meter long segment of pure LK99 you'd have a point. But the pieces in question are so small that even if the LK99 were just copper their resistance would be so low that experimental error in a casual test would be substantial compared to the resistance. Measuring the resistance would also be frustrated by the superconductive material being scattered in the mass separated by highly resistive impurities, making it easy to get a false negative.