AC mains and low voltage DC in same cable?

[rod]

Is it acceptable to use a spare core in a 4 core armoured cable to carry 12volts DC when other cores in the same cable are carrying AC mains?

 

[rod]

Many thanks for your replies all. This was to run LED markers on external posts using the spare capacity in the armoured power-cable, although perfectly capable from an insulation point of view it felt wrong mixing these voltages in the same cable, hence my post. This would have saved many transformers and resultant bulky housings at each location. I will not run the 12vdc.

 

[Wilko]

... I would still consider the implications of the LV crossing over to the ELV and the potential risks associated with it if it were to occur, fusing alone isn't always a sufficient, like I said I said before, if DC is generated and isolated from the LV side in its generation then this could potentially create a accident waiting to happen.

Yes I agree, the ELV may be very basic ("it's only 12V...") and probably shouldn't be in the same jb containment (?) which would likely happen if it was within a multicore swa. Perhaps for ELV control signals for a machine (that's being powered by the other cores of the swa, say) then it could be ok per 2nd option ?

 

[Deleted member 26818]

Over the years, I have been tasked with wiring a few Dali type lighting systems using 5 or 7 core cables.
3 or 4 cores for power and 2 cores for data.
The power is LV ac and the data ELV dc.
Cables have been either flex or SWA depending whether they are internal or external.
None of the cables have had internal screening between the LV and ELV.

 

[davesparks]

Over the years, I have been tasked with wiring a few Dali type lighting systems using 5 or 7 core cables.
3 or 4 cores for power and 2 cores for data.
The power is LV ac and the data ELV dc.
Cables have been either flex or SWA depending whether they are internal or external.
None of the cables have had internal screening between the LV and ELV.

Yes, I have done similar and as far as I know this is perfectly compliant (conductors insulated to highest voltage)
And presumably the ELV data/control circuitry is designed such that it immune to 'noise' or other interference from the AC.

 

[Lucien Nunes]

One other consideration is the method of protection of the two circuits. The LV circuit is likely to be TN mains relying on ADS, while the ELV circuit might be expected to be electrically separated, perhaps SELV, and requiring more than simple separation at the terminations. A fault from the LV to the ELV such as tracking between terminals in a JB might not be detected and might persist. Within the cable, the ELV circuit is still insulated up to LV from the outside, so it doesn't create a shock risk there, but unless all the ELV fittings are insulated for LV they would pose a shock risk.

Re. sending ethernet over mains cables, this is a different scenario because it's not an ethernet 'circuit' and a power circuit sharing a cable, it's just the power circuit with the data modulated onto it electronically. The sender and receiver provide the electrical separation and the modulation scheme is designed to be tolerant of interference. You would not get the ethernet physical layer itself to go very far down SWA!

Data transmission systems designed to be integrated with LV AC power are usually differential, e.g. DALI / DMX /RS485 or current-driven if analogue e.g. 4-20mA which makes them particularly resistant to interference. Twisted-pair ethernet is also differential but it goes so much faster that the cable's inter-pair crosstalk is much more of a limiting factor than interference pickup from other circuits.

 

[tom porter]

How will you get induced voltage on the dc circuit if the cores are insulated to the highest voltage present i.e. 230volts?

do you not know what an induced voltage is?

 

[redbarnman]

What this discussion shows is how you can write rules that still don't make a requirement clear and unambiguous. An extension of this would mean no ELV (12 or 24v AC/DC) or signal cables in the same trunking or conduit as LV (mains). It may be undesirable because induction will put a small "50Hz hum" signal on the ELV and when mains switching of inductive elements takes place e.g washing machine motors or a resistive spike as an elecric kettle is turned on a small signal spike will be induced in the ELV circuit. Induction effects can be minimised by using twisted pair as in CAT5 etc. Seems to me there is no issue with using LNE and DC1 and DC2 in a 5 core cable so long as you can deal with the induced voltages within the ELV equipment.