Emergency Stop implementation

[Andy Dingley]

Most RCDs don't have a convenient earth terminated on them, so it's easier to do it by unbalancing the phase & neutral as they run through the current transformer. The detector circuit itself doesn't care how you do it.

Where you have a DC-sensing fault detection circuit (most RCBOs) it's not possible to do this simple imbalance, so they use the alternative of providing a simulated fault current to earth. Hence the need for a low-current earth connection on this type, usually as a small flying lead.

I'm not up to date on the legislative state of things, nor on the requirements between RCD / RCBO and AC / DC sensing. Pointers would be welcome.

[Harry Bloomfield]

Al wrote on 15/03/2009 :

Turning a motor or light on puts a resistor of sorts across the supply, as does pressing the RCD test button - the difference is?

On that we can agree.

[Dave Osborne]

The difference is a "resistor of sorts" between live and earth (RCD test button) as opposed to between live and neutral (normal load).

[John Rumm]

That sounds unlikely. Do you have a reference?

[Dave Osborne]

My bad, that should read 'The difference is a "resistor of sorts" between live and earth (deliberate remote tripping of an RCD)...'

[Andy Dingley]

The difference is that both connections to the load pass through the RCD's sensing current transformer (in a direction such that one cancels out the other and the nett current is zero), whilst the test load is only passed once through the sensor and so it represents an unbalanced current.

Current operated RCDs don't detect the "leakage current" per se (that would be difficult from a remote location), instead they detect an imbalance in Live and Neutral currents, then assume that any difference is a leakage to earth.

In fairness to Al, he's also talking against the variant where the test current is fed to earth, rather than this unbalanced current. My own sloppy wording might have sent the thread in that direction.