9 hours ago, Craig Sawyers said:

Simple opposing magnets are not stable. The only thing that is stopping this instability is the jewelled bearings - but because of the magnetic instability there is more force on the jewelled bearings to apply a centering force that might be naïvely be expected.

I guess that is why the Castagna model A was pretty much the only tone arm to use the magnetic principle. Although at around the same era there was a design and prototype for a parallel tracking arm the used opposing magnets in a saddle arrangement; but that never made it to production.

Now it is now possible to levitate stably. The idea being the the repelling magnets are stabilised actively using a set of coils. Around 8cm diameter (so not small) and because of the active control a potential source of electrical interference near to the cartridge.

Then there is Meissner effect levitation. That is not entirely daft with high temperature superconductors. Of course it needs the superconductor to be cold enough, so a source of liquid nitrogen is needed. I've done exactly that as a table top experiment when my kids were small, and I was working at Oxford Instruments, so had ready access to chunks HTc superconductor, and a small expanded polystyrene pot for the LN2.

https://oswaldsmillaudio.com/schroder_model_2

For another magnetic (ish) tonearm

I often wonder if the simplicity and negligible bearing friction of string or unipivot tonearms compared to more complicated gimbals is worth more than their tendency to wobble about any direction they feel like doing.