I'm a bit sheepish about mentioning this, but the sources I've researched suggest that motor lead length affects the VFD performance.
If the circuit is sensitive enough on the output stage to have that as a consideration, I kept my output leads to just over a meter, and went without filtering, for now. My solution for shielding was using a separate shielding braid salvaged from a 70s cabinet disk drive, it happened to be about the same length. -- I still have the 11 platter removable cartridge (!!)
At these low frequencies, even a steel open braid that is available by the foot would sufffice, The impedance at these frequencies and their harmonics would be easily be kept very low.
I don't know why you would be sheepish about saying any of that.
My "opinions" of your comment go as follows.
To preserve the sharp edges that are output by a VFD, short is way better than long.
Any shielding is better than no shielding.
Although I have not measured it, the frequencies involved are probably much higher than we think. Any fast square wave carrier will probably show a gigahertz component. But you are right, the main fundamental frequency of the basic carrier is in the audio frequency range so even speaker wire would suffice.
For the time being, I have used plain old Romex 3-wire (plus ground) home wiring on both mills my VFD has been attached to. Other than the scream in my Bluetooth hearing aides, it seems to work just fine.
Now we get to the more complicated discussion. To maintain the sharp duty cycle character if the VFD output to the motor and supress Radio frequency noise, the main VFD manufacturers all highly recommend individually shielded, 3 wire, balanced cable. Not only that, but they also recommend shielded grounding at both ends. The video I posted earlier is a good reference.
However, I also note that most of their recommendations involve dealing with very long cable connections. Very little that I have found deals with short cable. Generally, they simply say keep the cable as short as possible.
My existing romex connection works. However, if I can find good VFD rated cable without spending a fortune (about $10/ft or less) I will use it. If I don't find any, I probably won't bother with anything in between.
Unless I find compelling information to the contrary, I will not use ferrite rings. While the VFD will work with them, and while they will reduce ringing, I believe they are probably a step in the wrong direction because they also dampen the desirable sharp edge duty cycle component of the VFD output.
Anyway, I am of the overall opinion that rf from the output cables isn't nearly as big an issue as the noise on the input wiring that might be fed back into other equipment in the shop and even the house. So my first priority is plain old RL filtering on the input. Anyone using a phase Converter or input transformer probably doesn't even need to worry about it at all because either of these options will totally kill any noise of this kind.
