Inductive cause of leaks to earth

[JohnW2]

In the below extracts from another thread, it can be seen that it has been suggested that 'inductive coupling to earth' can create leakages to earth which can resultant in 'residual currents' that could trip an RCD.

As you can see below, I don't really get this, and therefore probably need some education - for the reasons I explain in the last of the below snippets. Thanks!

Kind Regards, John

Relevant bit from the very end of that “tech info”pdf:
RCD’s trip with an earth leakage of 30mA which is sufficiently low enough to protect against electric shock, and is the industry standard size of RCD. But what is unusual about sunbeds is that the choke (ballast) inside that supplies each lamp, by its very nature, has a natural earth leakage of around 0.75mA.. This is inherent in its design, as a copper coil surrounding a soft iron core. Hence a 40 tube sunbed can easily have a total earth leakage of 30mA, and a 50 tube sunbed almost certainly would.

I have to say that's a new one on me. Capacitive leakage to earth is straightforward enough, but I've not come across (and struggle a bit to understand) 'inductive leakage'. On the face of it, that concept would apply to any wirewound component (motor, transformer etc.) would suffer from the same issue. Can someone educate me?

I agree with AC capacitive and inductive linking can cause a leakage to earth even when with DC there would be no leakage ...

As I said, the 'inductive' case is a new one on me, so I need some education!

As I said, it's straightforward enough with capacitive coupling, since it is then possible for there to be a capacitive path to earth from the L of the supply - either through stray capacitance or, say, intentional filter capacitors etc. Even if (as will usually/often be the case) there is such capacitive coupling to earth from both L and N, the 'leakage current' through those paths will be much greater from the L than from the N (due to much larger pd across the capacitance), thereby leading to a net L-N current imbalance which, if large enough, will trip an RCD/RCBO.

However, in the case of the postulated 'inductive leakage', which an AC current flowing through an inductor will induce currents into any conductor within its field, I do not see how this can result in an imbalance of L-N currents in the supply to the inductor.

What am I missing?

 

[Jackrae]

I would surmise : The inductive coupling to 'earth' can only be via a magnetic field and as such does not constitute an electrical current flow from the 'primary' side of the choke device to the 'secondary' (earth). The 0.7ma referred to will be the current required to feed this magnetic coupling with no load on the choke and is not a "leakage" current.

 

[JohnW2]

I would surmise : The inductive coupling to 'earth' can only be via a magnetic field and as such does not constitute an electrical current flow from the 'primary' side of the choke device to the 'secondary' (earth). The 0.7ma referred to will be the current required to feed this magnetic coupling with no load on the choke and is not a "leakage" current.

... but why would that result in different currents in the L and N of what was supplying the choke?

Kind Regards, John

 

[AdrianUK]

I too an skeptical about "inductive" coupling to earth.

Unlike capacitive coupling, how can inductive coupling lead to a current imbalance? I think the original OP is getting confused about leakage flux which in not the same thing at all.

Another possible explanation: Circuits with multiple flourscent tubes often have capacitors to correct for the poor pf due to the choke.

Are these the real cause of the leakage to earth?

 

[Jackrae]

I still maintain that there is no actual current flow across a magnetic coupling (until someone advises me otherwise)
The whole concept of transformer based safety isolation is that there is no possibility of current flow between primary and secondary

Do we have sufficient facts to address the question of where the 0.7ma supposedly travels
Is there really a L/N imbalance.
Does the device have some form of N-E circuitry
What type of leakage tripping device is in use. (RCD or RCBO) If RCBO could it be inrush current
Is there a circuit diagram of the device to determine other possible causes of "current" imbalance. etc etc

 

[JohnW2]

I too an skeptical about "inductive" coupling to earth. Unlike capacitive coupling, how can inductive coupling lead to a current imbalance?

Exactly - that's what I don't understand, either - hence my questions.

I think the original OP is getting confused about leakage flux which in not the same thing at all.

I don't think it's just the OP. The material endecotp quoted was very clear in saying that it was because the 40 x 0.75 mA 'leaks' could add/multiply up to enough to trip a 30mA RCD that they were saying that a 100 mA one had to be used - so they obviously at least thought that they were talking about imbalance currents.

Another possible explanation: Circuits with multiple flourscent tubes often have capacitors to correct for the poor pf due to the choke. .... Are these the real cause of the leakage to earth?

Certainly a lot more credible. A total of about 0.4 μF would be needed to get 30mA (at 50 Hz), hence about half that might trip a "30 mA" device.

Kind Regards, John

 

[Jackrae]

But surely the PF correction is between L and N and therefore not liable to be the route for earth leakage.
More like they are using some form of imperfect insulating material which has a 'leakage' to earth

Suggesting a 100mA protection device on a piece of equipment used in intimate contact with the human body is grossly irresponsible.

 

[JohnW2]

I still maintain that there is no actual current flow across a magnetic coupling (until someone advises me otherwise). The whole concept of transformer based safety isolation is that there is no possibility of current flow between primary and secondary

Exactly - and that's my point. There is no way that current can flow selectively from the L supplying the primary to the 'secondary' from which it is electrically isolated.

Do we have sufficient facts to address the question of where the 0.7ma supposedly travels
Is there really a L/N imbalance.

Only the fact that the material quoted suggested that the total of all the 0.75 mA leaks could result in an30 mA RCD tripping, so they obviously thought that it was somehow flowing from the L of the supply to earth.

Does the device have some form of N-E circuitry

I'm not sure what you mean,or how it alters what I'm saying.

What type of leakage tripping device is in use. (RCD or RCBO) If RCBO could it be inrush current

As above, they were clearly talking about the summation of 0.75 mA per tube 'leaks' tripping an RCD - so obviously nothing to do with inrush currents.

Is there a circuit diagram of the device to determine other possible causes of "current" imbalance. etc etc

Not that I am aware of.

My view is that whoever wrote the material that endecotp quoted had no understanding and was writing total nonsense. It might well be that, as Adrian has suggested, the sunbed is capable of tripping a 30 mA RCD because of leakage current through capacitors - but that is totally different from what they were suggesting.

Kind Regards, John

 

[JohnW2]

But surely the PF correction is between L and N and therefore not liable to be the route for earth leakage.

True, but I suppose that there might be L-E 'filter' capacitors.

More like they are using some form of imperfect insulating material which has a 'leakage' to earth

Conceivably, but insulation so poor that it allowed 0.75 mA per tube leakage would be quite ridiculous.

Suggesting a 100mA protection device on a piece of equipment used in intimate contact with the human body is grossly irresponsible.

Quite so - and a point I repeatedly made in the other (now locked) thread.

One can but assume that they know that the sunbed can trip a 30 mA RCD, so the question is "why?" - given that the 'explanation' they have offered makes no sense (at least, not to me).

Kind Regards, john

 

[Harry Bloomfield]

True, but I suppose that there might be L-E 'filter' capacitors.
Conceivably, but insulation so poor that it allowed 0.75 mA per tube leakage would be quite ridiculous.
Quite so - and a point I repeatedly made in the other (now locked) thread.

One can but assume that they know that the sunbed can trip a 30 mA RCD, so the question is "why?" - given that the 'explanation' they have offered makes no sense (at least, not to me).

Kind Regards, john

The simple solution is to measure the current flowing in the earth conductor, using an ac milliamp meter in series with such a light fitting, transformer or etc..

 

[JohnW2]

The simple solution is to measure the current flowing in the earth conductor, using an ac milliamp meter in series with such a light fitting, transformer or etc..

The 'simple solution' to what? As I said ...

One can but assume that they know that the sunbed can trip a 30 mA RCD ...

... which, if true, would mean that the result of your measurement would be totally predictable, but that would not help with the 'actual question' - which, as I went on to say, is ..

... so the question is "why?" - given that the 'explanation' they have offered makes no sense (at least, not to me).

Kind Regards, John

 

[333rocky333]

Dont know the answer, but a bloke on ebay has sold over fifty 55uf sunbed capaciters and one bloke selling 80uf for sunbeds.

Also appears some sunbeds still use big old magnetic ballasts, i know when similar ballasts are used for son lighting, a 100w son lamp ballast would only need a 12uf capacitor across L and N of each fittings supply.

 

[SUNRAY]

The simple solution is to measure the current flowing in the earth conductor, using an ac milliamp meter in series with such a light fitting, transformer or etc..

Out of curiosity I've just measured the flou fittings which I use for portable work. Started before your post.

These are all very old, all certainly pre 1994 and most are a decent number of years before that. Total of 17 including 6x 40W 'quickstarts' (fat tubes only) containing 2 wound devices, 6x 30W 3ft and selection of 5x 40W . All meggared as >400MΩ (some >2000MΩ) and earth current measured 6 lamps at a time; 6X quickstarts = 0.9mA (for all 6) 30W=0.2mA/6 and the others & 0.4mA/5. Or 1.5mA for all 17 which equates to 3.5mA/40 which will hopefully run forever on a 30mA RCD. however there are 2 more; 4MΩ/1.3mA & 11MΩ/4.2mA. These last couple are now scrapped.

I do give them a quick PAT (just L&N-E meggar and earth cont'y) when deployed but never thought to measure earth leakage before.


EDIT: just twigged mine are ≤40W or under so potentially have to multiply by 5. ≥

 

[JohnW2]

Dont know the answer, but a bloke on ebay has sold over fifty 55uf sunbed capaciters and one bloke selling 80uf for sunbeds.

If you connected them L-E you ought to get a 'leakage' current of about 4A or 6.4A (4,000 and 6,400 mA) respectively !! However, as previously discussed, if they were connected L-N, that would not result in any 'current imbalance' for an RCD not notice.

Kind Regards, John

 

[Jackrae]

Out of curiosity I've just measured the flou fittings which I use for portable work. Started before your post.

These are all very old, all certainly pre 1994 and most are a decent number of years before that. Total of 17 including 6x 40W 'quickstarts' (fat tubes only) containing 2 wound devices, 6x 30W 3ft and selection of 5x 40W . All meggared as >400MΩ (some >2000MΩ) and earth current measured 6 lamps at a time; 6X quickstarts = 0.9mA (for all 6) 30W=0.2mA/6 and the others & 0.4mA/5. Or 1.5mA for all 17 which equates to 3.5mA/40 which will hopefully run forever on a 30mA RCD. however there are 2 more; 4MΩ/1.3mA & 11MΩ/4.2mA. These last couple are now scrapped.

I do give them a quick PAT (just L&N-E meggar and earth cont'y) when deployed but never thought to measure earth leakage before.


EDIT: just twigged mine are ≤40W or under so potentially have to multiply by 5. ≥

What equipment are you using to measure the earth currents and how close is the test equipment to the ballast chokes