7
\$\begingroup\$

I have a doubt regarding the earthing of an appliance and its separate isolation transformer. Consider the following simplified schematics; this is my first guess at which earth each chassis should be connected to: connect the appliance chassis to the local earth, not the supply earth. Is it correct? What is and why? I'm sure it has mostly to do with the tripping of the residual current device when a fault current occurs to enclosures, but I haven't been able to get to the bottom of it.

This is also a general question, as I need to make sure the isolation transformer can be removed (appliance directly connected to the supply) without rewiring anything: there are therefore 2 cases of which one (n°2) is totally generic.

schematic

simulate this circuit – Schematic created using CircuitLab

schematic

simulate this circuit

Context: France and Belgium.

\$\endgroup\$
5
  • \$\begingroup\$ What is the length of cabling at the secondary of the isolation transformer? Is any transformer at the load side (use of linear power supply)? \$\endgroup\$
    – GR Tech
    Commented Jun 5, 2015 at 18:35
  • \$\begingroup\$ I don't think so, the load is made of IT equipment mostly (computers, monitors, etc.) along with power resistors switched by relays. The length of cabling between the transformer and the appliance is around 2 meters. \$\endgroup\$ Commented Jun 5, 2015 at 23:26
  • 1
    \$\begingroup\$ Earthing has nothing to do with the residual current device - that will work with or without earthing as if just measures the current difference between supply and return. The earth is for your own protection: without an earthed chassis, it could become live without and incident until you touch it. The earth ensures a live chassis will 'drain', in turn causing a differential between supply and return, tripping the RCD before it trips because you touched it (hopefully). \$\endgroup\$
    – RJR
    Commented Jun 9, 2015 at 3:54
  • \$\begingroup\$ Yes, I realised that but it is a good reminder. For these reasons I need it in there, along with proper grounding. Care for an answer? \$\endgroup\$ Commented Jun 9, 2015 at 9:56
  • \$\begingroup\$ Depends on the purpose of the isolation transformer. Are you using it for testing, or to isolate the appliance from a voltage offset? \$\endgroup\$
    – bigjosh
    Commented Jun 10, 2015 at 14:39

2 Answers 2

2
+50
\$\begingroup\$

IMO, both your cases are incorrect.

  • the connection between a provider distribution grid as the supplier and a house (home/building) as the consumer typically looks like that (remember, at least one full, i.e. L+N, breaker should be installed after the feeder enters into the house and before the in-house wiring/distribution):

    enter image description here

  • by this, earthing the in-house apparatus should be like that (remember, you need to use a multi-pole [L+N] breaker before each critical apparatus, e.g. xformer or welder, instead of a simple one-pole [L] breaker if you cannon ensure the direct wiring, i.e. L to L, n to N, not L to N, will take place all the cases and times):

    enter image description here

  • and last, if you decide to place an insulating xformer (by some your reasons which are not clear from the question) that could be if the distance between the feeder input and the place is too long to the PE wire to act as a protection correct, its case (if any) should be earthed (grounded) locally as well as the new PE wire going to the apparatus in the new insulated domain (remember, the L+N breaker mean also rules here):

    enter image description here

Therefore, my advice for you is the next:

  • provide you apparatus with an electric plug wiring out L, N, and PE (connected to the case, if any and conductive) circuits;

  • provide you xformer with an electric socket compatible with the plug on the apparatus, with its PE connected to the locally grounded circuit---at the secondary side (winding), and with an electric plug or terminals wiring L and N, but not PE, compatible with the socket or wires on the in-house power wiring---at the primary side (winding).

Good luck.

P.S. Above is related to TN-C based distribution grids typical for most Europe, imo.

\$\endgroup\$
9
  • \$\begingroup\$ Thank you for this well-made answer (+1). In France (and Belgium I think), the earthing scheme is TT. Are my cases still incorrect for this scheme? How would you modify your answer for TT (I'll definitely grant you the bounty then)? \$\endgroup\$ Commented Jun 14, 2015 at 20:03
  • \$\begingroup\$ I am not familiar with TT close and now having no technical ability to drawing pictures (therefore your bounty is unreachable for me :-) but i like to give your next additional comments/advices: \$\endgroup\$
    – asndre
    Commented Jun 15, 2015 at 10:20
  • \$\begingroup\$ 1) if your system is single-phase TT, there is no third wire between the supplier and the consumer at all, i.e. only L and N wires exist between them. at both your pictures the third wire must be eliminated and each one-pole breaker must be changed to a two-pole breaker. \$\endgroup\$
    – asndre
    Commented Jun 15, 2015 at 10:23
  • \$\begingroup\$ 2) the only significant difference between TN-C(-S) and TT for you is where to tie N circuit to the local earth: before the breaker (as typical for TN-C) or after (as typical for TT) as it seems from the consumer (burden) point of view. \$\endgroup\$
    – asndre
    Commented Jun 15, 2015 at 10:27
  • \$\begingroup\$ Right, you seem to say it is correct for a TT scheme as long as the breaker is 2 pole. I'll keep the TNC scheme in mind in case it is the one implemented. Cheers \$\endgroup\$ Commented Jun 15, 2015 at 10:29
1
\$\begingroup\$

The isolation transformer has one of its output terminals earthed on your schematic This means that the power isn't really isolated at all BUT your earth leak breaker will work IF you keep the isolation transformer isolated by not grounding any of its output terminals AND want earth leak operation then you can ground the centre tap via a resistor If a center tap connection isn't conveniently available then two equal value resisters will achieve the desired result at the expense of more power draw The resisters should allow at least 30mA to flow if either line is faulted to ground so normal earth leakage breaker operation will occur

\$\endgroup\$
1
  • \$\begingroup\$ Good ideas (+1). The absence of full-stops at the end of your sentences makes the answer difficult to follow and sometimes ambiguous though; you should certainly add them for other readers. What about the main issue about which ground to connect to chassis? Could you copy my schematic and modify it to illustrate? \$\endgroup\$ Commented Jun 9, 2015 at 10:07

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.