What is the frequency range you anticipate using ? - the effects of inductance start to bite as the frequency goes up, and this means that you need to calculate for the skin effect on sizes where at or near 50Hz you would not bother.

In practice this may need to a slightly larger cable than you might have expected. Also note that the waveform from VFDs is far from sinusoidal, and there is more scope for interference with small signal circuits carrying video, voice or data. As such screened cable may be needed, and/or more attention than normal to segregation .

Mike.

200Hz at 133V output, which needs to go through a local distribution board to feed commando sockets (which are rated between 100-300Hz) to specialist power tools.  This is a correction to what I stated earlier about VFDs.

200Hz at 133V output, which needs to go through a local distribution board to feed commando sockets (which are rated between 100-300Hz) to specialist power tools.  This is a correction to what I stated earlier about VFDs.

OK at 50 Hz skin depth in copper is about 1cm, so any wire of diameter quite a bit less than that can be considered as much the same as DC.
300Hz skin depth in copper is more like 4mm, so the onset of having to think about it is at core diameters of half the size. (its a square root of frequency thing, so every time the frequency goes up by ten, the distance goes down by 3 and a bit.)

commando sockets suggests flex and up to about 4mm2 I'd use the normal 50Hz tables and not blink.
BUT how are earthing and ADS organized ?

Mike