I've played with pair of 3" ring magnets: it easily supports my 32 kg 4" thick platter placed below it. But the magnet field went up to 20 gauss measured at the record plane - to much! Three layers of mu-metal sheets did not help much. There is another inherent drawback: magnet repulsion is essentially unstable; every minor deviation from perfectly concentric magnet fields (unevitable) produces considerable side force applied on journal bearing, increasing bearing noise. So I finally abandoned this idea.
