Unfortunately too much speculation has created conclusions. Some misinformation noted and corrected.

One, earth ground has no relationship to how an ELCD, RDC works. That is completely about what safety ground does. Unfortunately some mistakenly call that earth ground. Safety ground means a fault to any equipment results in a tripped circuit breaker. So that humans are protected from shocks. That receptacle's safety ground cannot connect to earth ground. It must connect to a bus bar in the main breaker box. Otherwise it may not provide sufficient human protection. And so the two ground (even if interconnected elsewhere) are completely different - have different purposes.

Second, ELCD, etc detect when an incoming current is different from an outgoing current. That means some current must be going elsewhere; maybe harmfully through a human. So the ELCD/RCD cuts off electricity. It does not need any safety ground or earth ground to function / protect / be effective.

Third, ELCD/RCD is not tripped by a microseconds current from something like lightning. A fault current takes milliseconds. But that much longer current can be created by a 'follow-through' current. This is AC current s flowing because a microsecond transient has made other and temporary connections (ie plasma or arcs).

Tripping is not created by lightning. But what lightning created and must never create causes an ELCD/RCD to trip. Solution was always this simple. Do not give a lightning current any reason to be inside. Otherwise all appliances are at risk.

That current is inside because it found better paths to earth (sometimes destructively) via appliances. If that current connects to earth BEFORE entering, then nothing inside is conducting that microsecond current. No microsecond current creating a milliseconds follow-through current means no nuisance tripping.

Nuisance tripping is a misnomer. Since the problem is really crated by intentionally not earthing transients to earth ground (not safety ground) and BEFORE it can enter.

Four, an effective solution means every incoming wire must connect low impedance (ie less than 3 meters) to earth at the service entrance. TV cable and satellite dish would make that connection only with a hardwire. AC electric and phone wires must make that low impedance (ie less than 3 meter) connection via a 'whole house' protector.

If that current is not connected low impedance (ie hardwire has no sharp bends) to upgraded earth ground electrodes, then that current will be inside hunting for earth ground via interior appliances and wires. That can even cause nuisance tripping. Because a current that should not be inside is inside.

That one human mistake that creates nuisance tripping. Another can be electronics that are missing required line filters. Resulting high frequency noise can cause some ELCD/RCD circuits to become confused and to intermittently trip.

Do not cure symptoms. Sensitivity is not the problem. Currents that should not be anywhere inside are the problem and what must be addressed. Those currents must connect low impedance (ie hardwire has no splices) to earth ground electrodes BEFORE entering.

Even 1 ma can kill. Some nations require ELCB to be 5 milliamps. Even 100 milliamps is dangerous. Why then would anyone use a 300 ma ELCB (or why is it even available)?

This post has been edited by westom: Oct 22 2019, 05:24 AM

It is accomplished by putting one on every branch circuit.

Wire quality, receptacles, and plugs are not significant. Appliances can leak many tens of microamps. Or sometimes as much as hundred microamps. One RCD means it must not trip when all appliances are leaking 100 or 200 microamps.

Originally, RCDs were only required on bathroom and kitchen circuits.

Interesting his how the guy who developed it also demonstrated it. He had his daughter in a bathtub. And threw a powered radio in that bathtub. I had to reread the article many times before I finally understood (accepted) it. That was back when radios did not use transistors. The technology is that old.

This post has been edited by westom: Oct 22 2019, 04:59 PM

Current price for some RCDs is as low at 60 rm. Some that are installed in the breaker box (as a circuit breaker) can be 250 rm.

This post has been edited by westom: Oct 22 2019, 05:05 PM

Then we put numbers to it. A lightning strike nearby to a long wire antenna (designed to be most sensitive to E-M fields) can result in thousands of volts on that antenna wire. Then connect a one milliamp Neon glow lamp (ie NE-2) to that antenna lead. Voltage then drops to maybe 60 volts. Because the induce surge has high voltage if current does not flow. And has near zero voltage if even less than one milliamp flows.

Induced surges are made irrelevant by what is already inside every electrical device.

Lightning struck a lightning rod. That means maybe 20,000 amps was flowing down the lightning rod's hardwire to earth. Just 1.3 meters away, inside, was an IBM PC. It did not even blink. And either did any other office electronics. That was a major E-M field confronting electronics - that did not even cause a software program to flicker. Because the destructive power of nearby E-M fields is overhyped by speculation. And does not exist once we include relevant numbers.

A tree was struck by lightning. Some ten meters distant, a cow died. Wild speculation assumes the cow was killed by an induced surge. Of course not. Lightning is a connection from a cloud (maybe 3 km up) to earthborne charges (maybe 4 km distant). That path also went up the cows hind legs and down its fore legs. What only observation speculated was an induced surge, in reality, was a direct strike. Conclusions only from observation create junk science. Once the many facts (with numbers) are learned, then that observation results in a completely different conclusion.

Nearby lightning strikes are either direct strikes or do not do damage. Nearby strikes that would trip an ELCD/RCD must somehow create what trips it. Not a voltage. A current that exceeds 100 ma. Induced surges just do not have that current. But many assume it was a induced surge rather than discover it was actually a direct strike.

Furthermore an ELCD/RCD requires that current to flow for tens of milliseconds or longer. Lightning is a microseconds event. too short to trip it. Something completely different (ie a follow-through current) may be relevant. But that is not a nuisance trip. That is a problem that must be protected from. And that is something that should not happen if other solutions are properly implemented.

This post has been edited by westom: Oct 22 2019, 11:23 PM

And that current is microamps or less. Much of electricity flows outside the wire - the skin effect. Like the nearby lighting strike, it can create high voltages only when no current is flowing. ELCD/RCD is not about voltage. It is about current. That leakage may be in the microamp or nanoamp region. Well below what is normal leakage in electronics (powered off or on).


This post has been edited by westom: Oct 22 2019, 11:28 PM

Skin effect is why AC power wires cannot carry any more current if thicker. And why voltage on one wire can create a voltage (with almost no current) on another. Skin effect applies to all frequencies down to but not including zero hertz. The point - skin effect explains why one wire can induce a voltage on another. And create maybe nanoamps of current - virtually a zero current.

So what is relevant? Leakage from wires is totally irrelevant - once we put numbers to that speculation. Most leakages are from appliances. Which is why some RCDs do not trip for a massive (less than) 5 milliamps. And why an RCD of 100 milliamps is massively above that normal leakage currents.

Described were some reasons for 'nuisance tripping'. In each case, due to some actual defect. For 'in the wall' wires to create nuisance tripping, insulation of two wires must be so massively compromised that wire separation is well less than 1 mm.

There are simple techniques for finding that defect. But the usual suspects are being ignored. ELCB/RCDs are almost last on a list of possible suspects. Tripping is reporting a defect.

Again, due to not having required solutions, then lightning (that did not trip that ELCB) creates a follow through current. We know a follow through current existed because that current is long enough to trip an ELCB. Lightning it not. Time is another parameter that must not be ignored.

We also know that comm ports are usually damaged not due to lightning on the communication lines. It is most often due to lightning incoming on AC mains. Then outgoing via comm lines. Damage is often on the outgoing path. A fact that get ignored if a conclusion is made only using observation.

A direct strike to all household appliances can be lightning striking poles many blocks away. But again, it is not the lightning that trips the ELCB. And that follow through current does not exist if appropriate solutions were implemented.

20,000 amps is too short to trip an ELCB. But the resulting 'follow-through' current created by that voltage means a large current that is longitudinal mode. Those type of currents easily trip ELCB/RCDs. 'Follow-through current does not exist if appropriate solutions were implemented.

The point of skin effect is overlooked. You said wires leak current. That that leakage can even be measured by 20 volts on a meter. That 20 volts is not leakage currents. It is an induced voltage created because current flows outside of a wire - skin effect. Current associated with that 'skin effect' created voltage is probably nanoamperes. Completely irrelevant to any ELCB/RCD operation.

Wires are not the typical source of microamp leakage currents. Appliances are. Even that resulting current is so low that a 5 ma RCDs should not trip. So a 100 ma RCD is well above what would cause an ELCB/RCD trip.

A 100 ma trip is saying that some defect existed. One example of why is 'follow through' current. Correct the reason for that current rather than blame induced currents by adjacent L-N wires. Replacing a 100 ma RCD with a 300 ma one is only to cure symptoms. Instead, address the actual problem.

Skin effect is and because of E-M fields. Only pure DC stays inside copper wire. As frequency increase, more of the E-M field is outside that wire.. As noted, 2.4 Ghz has much more outside the wire. Skin effect or what induces on adjacent wires exists with any frequency as long as it is above zero. Skin effect is why AC electricity is typically only inside (if I remember the number) 9 mm inside a wire.

So yes, 20 volts is possible if two wires are together long enough and no current is conducted on the 'induced' wire. But once that wire starts conducting nanoamps, then that voltage drops to near zero. Nanoamps or even microamps are no where near relevant to a 100 ma ELCB/RCD. Numbers say move on to other suspects.

So what causes over 100 ma to trip that ELCB/RCD? Not just any current. It must be a longitudinal current. And it must be long enough for an ELCB/RCD to respond. A likely suspect is 'follow-through' current. It exists because that lightning current was all but invited inside. Creating temporary and conductive paths that permit a follow-through current (a longitudinal mode current) to exist long enough and strong enough to trip that ELCB/RCD.

Obviously replacing that circuit breaker or installing one with a higher rating would neither avert another trip or even address the problem. But again, this is only one suspect. A likely one that has been observed and eliminated elsewhere. And one not solved by so many other suggestions.

This post has been edited by westom: Oct 25 2019, 12:07 PM