alt Hacker NewsI do this stuff for a living, and have an enormous amount of experience with it. There are worse things than ground loops, and USB cables with broken shields are one of them. Braids in sensitive applications are essentially never grounded at only one end.
Ferrites do nothing for low-frequency impulse noise, which is what this type of issue usually comes down to. They are there to pass FCC/CE tests, and that's it. Optically-isolated cables can indeed be a good solution, but not many of them support USB 2.0 HS unfortunately (which my stuff needs). Might be the right answer here.
I got noise out of 5V TTL encoder signals with clamp-around ferrites around the motor power leads. The encoder input was picking up the servomotor power.
The big problem here is that you can't do much about the source. A plasma arc started with audio-frequency AC (10KHz?) is going to emit RF. Not clear what the waveform looks like. If it's a square wave, which it probably is, there will be higher harmonics. Fortunately, it's only AC only during start; the cutter shifts to DC once the arc has been struck. Otherwise those things would be a huge interference headache whenever they were on.