RCD wiki + Common Misconceptions about RCDs

[sparkticus]

I'm also interested to see the 3.9k resistor, which I presume was attached to the test button - which implies a 'test' current of about 60 mA.

Yes, the resister connects (when test button operated) line and neutral via the PCB. For a moment I thought that something on the PCB may have provided an additional (series) resistance but although the the resister connects to Neutral via the PCB(before the toroid), I am able to measure 3.85K across line and neutral. I wonder why they picked 60mA perhaps simply because it is twice the 30mA threshold and for that reason should always trip the device.

 

[DetlefSchmitz]

I wonder why they picked 60mA perhaps simply because it is twice the 30mA threshold and for that reason should always trip the device.

...even if there is already a 29mA imbalance in the other direction.

 

[sparkticus]

I wonder why they picked 60mA perhaps simply because it is twice the 30mA threshold and for that reason should always trip the device.

...even if there is already a 29mA imbalance in the other direction.

That's what I'm looking at at the moment. I'm trying to see what effect N>E current (from N-E PD derived from a TN-S supply) has upon the trip threshold. I'm using two potentiometers but it is tricky setting up the N>E potential without confusing the results.

 

[JohnW2]

I wonder why they picked 60mA perhaps simply because it is twice the 30mA threshold and for that reason should always trip the device.

...even if there is already a 29mA imbalance in the other direction.

Good point - that's probably the reason.

Kind Regards, John.

 

[sparkticus]

I wonder why they picked 60mA perhaps simply because it is twice the 30mA threshold and for that reason should always trip the device.

...even if there is already a 29mA imbalance in the other direction.

Ah yes, I see your point. Up to 29mA RCD output neutral to line input side.

 

[JohnW2]

That's what I'm looking at at the moment. I'm trying to see what effect N>E current (from N-E PD derived from a TN-S supply) has upon the trip threshold. I'm using two potentiometers but it is tricky setting up the N>E potential without confusing the results.

Provided you were careful to get phases right ( ) (might need to check with a scope), I would have thought that probably the 'least potentially confusing' way to created the N-E PD for your experiments would be to use, say, a 6V transformer, with it's secondary connected between N & E of your test rig.

If you weren't already doing it, hence saving me the effort, I'd be doing the very same experiments this weekend!

Kind Regards, John.

 

[sparkticus]

I would have thought that probably the 'least potentially confusing' way to created the N-E PD for your experiments would be to use, say, a 6V transformer, with it's secondary connected between N & E of your test rig.

If you weren't already doing it, hence saving me the effort, I'd be doing the very same experiments this weekend!

Yep, that's essentially what I have got. I only have a 12VAC transformer but I can still get the correct current with a fixed resister and a pot (to fine tune) in series. It's amazing how easily RCDs trip. Not unexpected of course but a bit like trying to keep a set of scales balanced while riding one of those motorised bucking-bronco's

 

[JohnW2]

Yep, that's essentially what I have got. I only have a 12VAC transformer but I can still get the correct current with a fixed resister and a pot (to fine tune) in series.

That sounds fine - the actual voltage is obviously unimportant. I presume you have managed to convince yourself that the N-E voltagesource is in phase with the L-N (or L-E) voltage?

It's amazing how easily RCDs trip. Not unexpected of course but a bit like trying to keep a set of scales balanced while riding one of those motorised bucking-bronco's

Yes, I can believe that - a bit like trying to balance a high-gain differential amplifier! It's just a matter of patience!

Kind Regards, John.

 

[sparkticus]

Yep, that's essentially what I have got. I only have a 12VAC transformer but I can still get the correct current with a fixed resister and a pot (to fine tune) in series.

That sounds fine - the actual voltage is obviously unimportant. I presume you have managed to convince yourself that the N-E voltagesource is in phase with the L-N (or L-E) voltage?

It's amazing how easily RCDs trip. Not unexpected of course but a bit like trying to keep a set of scales balanced while riding one of those motorised bucking-bronco's

Yes, I can believe that - a bit like trying to balance a high-gain differential amplifier! It's just a matter of patience!

Apologies for the delay, a few domestic tasks (assigned to be by my daughter by executive decision and possibly conspiracy) got in the way. Should have results tomorrow.

 

[JohnW2]

Apologies for the delay, a few domestic tasks (assigned to be by my daughter by executive decision and possibly conspiracy) got in the way. Should have results tomorrow.

Thanks - no rush. Accuracy is more important than speed

Kind Regards, John.

 

[SimonH2]

Loss of neutral on a 3 phase main in the street very, very rarely leads to reports of shocks etc. Lots of other effects but rarely that! These faults are fairly rare and usually only occur on one type of underground cable that is not in use in all the DNOs.

Or when some scrote nicks the N-E link at the substation ?

One other misconception that's sort of touched upon in a couple of places is that of discrimination. Ie, some people believe that if you have (say) a 100mA RCD followed by a 30mA RCD, then the 30mA device will trip but the 100mA one won't.

This is hinted at in the section on TT systems and the use of a delayed type, and also relevant to the section where it says that RCDs do not limit the fault current.

2...I've yet to find a good explanation of the inner workings of an RCBO, but is is often said that the 'residual current' part of it is 'electronic', rather than electromagnetic as in the RCD. If that's true, what does that actually mean (I rather suspect that there must be some wound components involved), and how does it cope with the phase issues which we've realised are crucial to these recent discussions? Indeed, do the conclusions that we have eventually come to (about RCDs) actually apply to an RCBO?

There is no difference between the RCD part of an RCBO and a standalone RCD - well none that are specific to the different type of devices, there may well be implementation differences. In both cases, the sensing system is typically a small toroidal transformer which detects the difference between line and neutral currents. In principal it would be possible to use the current in the sense winding to directly operate a trip, but in practice, it's easier to make a small electronic circuit to allow precision in both tripping level and time - a purely electromechanical system would trip faster under higher fault currents (more force = higher acceleration on parts) and a reliable, precise mechanical delay is not easy to achieve.

 

[JohnW2]

One other misconception that's sort of touched upon in a couple of places is that of discrimination. Ie, some people believe that if you have (say) a 100mA RCD followed by a 30mA RCD, then the 30mA device will trip but the 100mA one won't.
This is hinted at in the section on TT systems and the use of a delayed type, and also relevant to the section where it says that RCDs do not limit the fault current.

Good points. Thanks.

2...I've yet to find a good explanation of the inner workings of an RCBO, but is is often said that the 'residual current' part of it is 'electronic', rather than electromagnetic as in the RCD. If that's true, what does that actually mean ....

There is no difference between the RCD part of an RCBO and a standalone RCD - well none that are specific to the different type of devices, there may well be implementation differences. In both cases, the sensing system is typically a small toroidal transformer which detects the difference between line and neutral currents. In principal it would be possible to use the current in the sense winding to directly operate a trip, but in practice, it's easier to make a small electronic circuit to allow precision in both tripping level and time - a purely electromechanical system would trip faster under higher fault currents (more force = higher acceleration on parts) and a reliable, precise mechanical delay is not easy to achieve.

Many thanks for clarifying; that all makes sense. In that case, is it yet another misconception ('urban myth') that RCBOs are specifically more suspectibe to damge from IR testing? Is it perhaps simply that, unlike RCBOs, RCDs would rarely be 'in circuit' (unless contained in sockets or other accessories) during IR testing?

Kind Regards, John.

 

[SimonH2]

In that case, is it yet another misconception ('urban myth') that RCBOs are specifically more suspectibe to damge from IR testing? Is it perhaps simply that, unlike RCBOs, RCDs would rarely be 'in circuit' (unless contained in sockets or other accessories) during IR testing?

Dunno, but apart from the specific case you've mentioned, there's no reason one should be more or less susceptible than the other. There may be a slight issue in that a 2 pole RCD probably has a little more room in the case for electronics, but that that's going to be a very small factor IMO.

Being able to knock off all the MCBs and not expose the RCD to the IR test would be an explanation for the difference in perception.

 

[securespark]

Looks a bit like I took it while galloping past on a horse...

On your way to another job, were you?

 

[EFLImpudence]

RCBOs do have a warning label against IR testing them.

RCCBs do not.

I have noticed that the tripping times of RCBOs are remarkably constant and similar.

I had put this down to being 'less mechanical' (than RCCBs) because of what I have heard.