Sensitivity in RCDs

[bernardgreen]

This is a theory that appears to fit with experience

Early RCDs did not have electronics and relied totally on the electrical energy created in the sensor's output winding to convert to mechanical energy adequate to operate the mechanical trip mechanism. The inertia of the system absorbed the energy created by transient impulses of un-balance between Live and Neutral currents.

Later RCDs used a simpler sensor with less windings ( Live and Neutral are not wound around the toroid core but pass straight through as a half turn ) thus less output which required electronics to measure the reduced output from the sensor. When the sensor output indicates an un-balance greater than 30 mA the electronics switch on an output to operate the trip mechanism.

The early sensors had two or three turns of the Live and Neutral leads around the toroid which was a much higher impedance to high frequency transients so the sensor output from transients and spikes would be far lower than they are in a modern half turn, low impedance sensor.

It is also possible that the electronics in a modern RCD will be saturated by a transient pulse from the sensor. This could be from a 1 amp unbalance that last less than a couple of micro-seconds. Such an unbalance could be created by loads being switched in circuits with considerable stray capacity between Neutral and Earth.

 

[ericmark]

I was going to direct you to a report on the Moeller X-Pole but it's gone. From memory they went to great lengths to prevent erroneous tripping which even included a LED warning light which changed colours giving a warning to when limits were being approached.

The problem is we really have no idea as to what has gone into the design. Before amendment 1 we could even fit auto resetting RCD's which clearly were rather complex. (Also expensive around £350 each)

It would seem today we rely on electronics in the old days a RCD was 4 module wide today they are normally only 2 module wide and RCBO are one module wide.

My mothers RBCO's have never tripped except for when using the RCD tester my old 4 module wide RCD's have bouts ever round and again of tripping then will settle again for maybe a year.

So yes there could be faults in the modern RCD which could vary make to make but how we could test to see if there are faults is another question.

In days gone by I made a crystal set radio and thought I understood how it worked today there is no way I am going to build my own DAB radio we just have to buy it an hope it works. Same with modern RCD.

 

[JohnW2]

Early RCDs did not have electronics and relied totally on the electrical energy created in the sensor's output winding to convert to mechanical energy adequate to operate the mechanical trip mechanism. The inertia of the system absorbed the energy created by transient impulses of un-balance between Live and Neutral currents.

True. I imagine that one downside of that approach was that it was probably impossible to achieve a precise and repeatable IΔn ('trip current'), since there would inevitably be some variability of the force (hence imbalance current) required to operate the mechanical trip mechanism.

Later RCDs used a simpler sensor with less windings ( Live and Neutral are not wound around the toroid core but pass straight through as a half turn ) thus less output which required electronics to measure the reduced output from the sensor. When the sensor output indicates an un-balance greater than 30 mA the electronics switch on an output to operate the trip mechanism.

Indeed. In contrast to the above, that ought to allow more precise control over the IΔn - the sensing of imbalance could be very precise and the 'output of the electronics' can provide an 'overkill' current to operate the trip mechanism, thereby largely overcoming any variability in that mechanism.

The early sensors had two or three turns of the Live and Neutral leads around the toroid which was a much higher impedance to high frequency transients so the sensor output from transients and spikes would be far lower than they are in a modern half turn, low impedance sensor.

True - but one would imagine/hope that there would be at least some attempt to filter out high frequency components prior to the electronics.

It is also possible that the electronics in a modern RCD will be saturated by a transient pulse from the sensor. This could be from a 1 amp unbalance that last less than a couple of micro-seconds. Such an unbalance could be created by loads being switched in circuits with considerable stray capacity between Neutral and Earth.

That's also possible but (a) as above, one would hope that some filtering would be present and (b) without a lot of detailed knowledge about the device, one could but speculate about what the consequences of this might be in a particular case.

Kind Regards, John

 

[EFLImpudence]

Am I being pedantic?

The outcome may be the same but I am always uncomfortable when people say RCDs detect the imbalance between L and N.

They detect and work by the residual current caused by this imbalance.

Is there a difference?

 

[JohnW2]

Am I being pedantic? The outcome may be the same but I am always uncomfortable when people say RCDs detect the imbalance between L and N. They detect and work by the residual current caused by this imbalance. Is there a difference?

I don't think there is a difference so I suppose that, yes, you are being a little 'pedantic'

"Residual current" is (L current minus N current), or vice versa, which is non-zero if L current and N current are different. "Imbalance between L and N (currents)" means that L and N currents are different, so that the 'residual current' is non-zero. Is there any difference? ... or are you perhaps merely quibbling about the fact that when people talk about "L-N imbalance" they don't usually explicitly say that they are talking about an imbalance (difference) between L and N currents?

Kind Regards, John

 

[EFLImpudence]

Not exactly.

It's just that the RCD sensor (the part which triggers the device) only monitors the residual current induced and not the actual current in the conductors.

Imbalance in the conductors could just as easily be termed 'current missing from the Neutral'.
Isn't it always L-N, not vice-versa?


An earth leakage clamp meter placed around the L and N wouldn't be said to be measuring 'imbalance'.

Just me, perhaps.

 

[JohnW2]

Not exactly. It's just that the RCD sensor (the part which triggers the device) only monitors the residual current induced and not the actual current in the conductors.

It does, but since that 'residual current' is the difference between the current in the two conductors, it surely comes down to the same thing, doesn't it. It really is just 'semantic'.

Imbalance in the conductors could just as easily be termed 'current missing from the Neutral'.

It could, but we don't generally use that term!

Isn't it always L-N, not vice-versa?

In practice, virtually inevitably. In theory a 'fault' (more likely a wiring error) which resulted in a connection between the L of the circuit and the L of another circuit not protected by the same RCD could result in a 'vice versa' situation (L current in the circuit less than N current - or, per your alternative terminology "current missing from the Line"!). However, one would have to struggle a bit to think of how such a 'fault' could arise (I suppose it could arise, particularly due to a wiring error, when there are Ls from more than one circuit in the same place, such as a multi-gang light switch - I suppose "a borrowed L")

An earth leakage clamp meter placed around the L and N wouldn't be said to be measuring 'imbalance'.

Well, maybe that language would not normally be used, but it could be - L-N difference/imbalance is precisely what the meter would be measuring.

Just me, perhaps.

Perhaps

Kind Regards, John