The cable must not be earthed at both ends, if the charging point has a different earthing arrangement due to PME (own TT for example) ... otherwise, you defeat the object of having TT to get round the PME issue.

Otherwise, earthing at both ends is usually better from an EMC point of view, but you will need to consider:

1. the effect of "defeating" an open-PEN detection device if one is fitted.
2. current-carrying capacity of the armour, as it will share any fault current of the mains power cable. An adiabatic check may be necessary. dependent on current sharing.

In all cases, SPDs may be necessary for the data cable in accordance with BS EN 50174-3, and may be required for the charging point itself, because of LPZ concept. SPDs almost certainly necessary for both power and data if the charging point has its own TT earthing arrangement.

From an electrical safety point of view could shielded cable create more issues than it resolves.

I suspect Fibre Optic may be the real answer, but that is not going to happen!

This is not shielded. It's armoured, and that is necessary to comply with BS 7671 unless you install the cable in a duct with equivalent mechanical protection (whatever that means ... but let's assume it's 750N twin-walled underground conduit ducting, for example)

So, not earthing it is a safety non-compliance, unless it's in suitable underground conduit.

Good bonding and short cable distances may remove the need for SPDs ... from an electrical safety point of view, if we think that's an issue, then, as normal Cat5/Cat6 etc. are internal cabling only, we might need to think about isolation barriers of internal-grade Ethernet equipment without the bonding?

Sparkingchip said:

I suspect Fibre Optic may be the real answer, but that is not going to happen!

Cable cost probably isn't prohibitive, but having the right splicing/terminating kit and skills might be a barrier.

Moving forward, as more so-called "smart" devices appear, perhaps that is they way we will all need to go anyway.

I was actually thinking of this armoured and shielded CAT6 cable

Cat6 FTP SWA 20m Cable (cablemonkey.co.uk)

The only time I have run CAT6 SWA was to a set of electric gates on the end of a long driveway, the groundworkers dug the trench then I ran the power SWA into IP rated double socket in a meter box built into the back of the wall by the gate along with the CAT6 SWA which I ran from a galvanised KO box inside the house earthed with a brass gland, with an ordinary CAT 6 from there to by the front door leaving the electric gate guy and the satellite TV guy who was doing all the networking in the house to finish it all off.

I doubt I am going to be able to duck out of completing everything in the future, leaving things to others to finish off.

Sparkingchip said:

I was actually thinking of this armoured and shielded CAT6 cable

From an EMC point of view, the shielding improves things as it's a lower impedance than the armour, and provides almost 100 % "coverage" of the signal conductors ... but the same rules go for the shielding as the earthing/bonding of the SWA. Certainly, never earth the shielding at both ends unless you earth the SWA at both ends.

Whether it's best to have screened data cable, or unscreened, depends on how noisy the environment is for radiated disturbances that might affect the cable.

The cable must not be earthed at both ends, if the charging point has a different earthing arrangement due to PME (own TT for example) ... otherwise, you defeat the object of having TT to get round the PME issue.

Don't connect the armour to TN at one and and TT at the other certainly. But presumably you could still connect both ends of the armour to the same (TN) earthing system at both ends - e.g. mains SWA (+G/Y core if used) to data SWA (+shield, if used) at the charge point, before the gap to the TT earthing system (all enclosed in an insulating enclosure).

   - Andy.

could and from the EMC point of view, usually should - but be careful that you are not asking a wispy braid to carry large fault currents or diverted neutral related offsets... That is another can of worms.

Mike.