'Leak' detection for RCD trip troubleshooting

[JohnW2]

In This Thread ...

Given the extent of difficulties (and time consumed) seemingly often involved in detecting the source of nuisance RCD trips, I'm rather surprised that we do not see (or do not see being used) 'leakage' monitors - or, at least, detectors. I'm sure, for example, that it would be easy enough to produce a relatively cheap plug-in device with a few LEDs which indicated whether L-N imbalance has exceeded certain levels (say 1, 5, 10, 20, 30, 50mA) during the period it had been in circuit. If the OP's electrician had such a device, a definitive answer about the appliance concerned could be achieved very easily. Do any such devices exist?

I would agree a residual current monitor (RCM) with recording would be a good tool but I also see some problems. First it would need to be on a non protected circuit and second would need a chart recorder not an every x seconds and join dots like used with current and voltage.

I don't see why it would have to be on a non-protected circuit. On the contrary, one would normally want to use it on an RCD-protected circuit, to ascertain whether a particular appliance had resulted in an RCD trip. Nor (in its simple form) would it need any continuous recording such as you describe. As I said before, it would be adequate to have a relatively cheap plug-in device with a few LEDs which indicated whether L-N imbalance has exceeded certain levels (say 1, 5, 10, 20, 30, 50mA) during the period it had been in circuit.

To be useful on an RCD-protected circuit, it would obviously have to either have a back-up battery or a non-volatile memory, but that's no problem. Such a device could probably quite easily be produced by poaching from RCD technology and electronics, and I'm sure could be fitted into a plug-in unit, and need not be particularly expensive. A more sophisticated version might store 'continuous' results in semiconductor memory (as you say, the sampling frequency would need to be pretty high), but that would really not be necessary (or necessarily useful) for simple 'culprit identification' purposes.

Kind Regards, John

 

[mfarrow]

Just make up a lead with a 10mA RCD in it. No appliance should ever leak that amount to earth anyway.

The problem with monitoring/measuring current is that the smaller the resolution you want to go to, the most expensive the kit. I wouldn't think anyone would successfully create a DIY version of what you want to build John without some trial and error. The concept is sound though.

 

[EFLImpudence]

Do they not exist already?

An earth leakage clamp meter can 'hold' the highest reading so all you would need is the batteries to last however long is necessary or to be replaced with an external larger one.

They are rather expensive - presumably because of the limited demand.

It is doubtful you would want such test equipment on a circuit which is not protected by an RCD.

A circuit which is protected by an RCD will obviously be disconnected when the appropriate leakage level is reached.
However, were the clamp meter measuring only the fridge circuit which had been noted to be leaking, say, 3mA when observed but after the RCD has tripped shows a high reading of, for example, 20mA it would be quite conclusive.

Whether any of us wish to leave our (expensive) equipment, or even several items, in a customer's property is another matter preferring to leave the suspected appliances disconnected one at a time to see what happens as now.

 

[JohnW2]

Just make up a lead with a 10mA RCD in it. No appliance should ever leak that amount to earth anyway.

Yes, that would work - it's basically just a 'one level' version of what I was 'thinking aloud' about.

The problem with monitoring/measuring current is that the smaller the resolution you want to go to, the most expensive the kit.

For the sort of purpose we're talking about, one wouldn't need high resolution.

I wouldn't think anyone would successfully create a DIY version of what you want to build John without some trial and error. The concept is sound though.

I wasn't really talking about a 'DIY version'. However, I'm quite sure that I could produce a device such as I described, using 'off the shelf' components (and maybe a sense toroid pinched from an RCD!) which would, in theory, be developable for commercialisation.

Kind Regards, John

 

[JohnW2]

Do they not exist already?

I don't know - that's precisely what I asked

An earth leakage clamp meter can 'hold' the highest reading so all you would need is the batteries to last however long is necessary or to be replaced with an external larger one.

I didn't even know that they existed - but, yes, it sounds as if they could probably be used (or adapted to be used) for the purpose. What exactly are they - just a low current clamp meter for a single conductor, or do they (like an RCD) clamp over an L/N pair and detect an imbalance? If not the later, do I take it that you'd use it on the CPC (which you would obviously have to 'expose' somehow). That obviously would not work if it were, say, a washing machine which was leaking current to earth primarily through pipework.

It is doubtful you would want such test equipment on a circuit which is not protected by an RCD.

Indeed - as I wrote to eric above.

However, were the clamp meter measuring only the fridge circuit which had been noted to be leaking, say, 3mA when observed but after the RCD has tripped shows a high reading of, for example, 20mA it would be quite conclusive.

Indeed.

Whether any of us wish to leave our (expensive) equipment, or even several items, in a customer's property is another matter ...

A fair point - although what I've suggested could theoretically be fairly cheap (and you could always charge a 'deposit' if you wished!).

... preferring to leave the suspected appliances disconnected one at a time to see what happens as now.

Again, often true - but if one is, say, getting nuisance trips every few days (or less frequently) and a freezer is one of the suspect appliances, I doubt that your customers would thank you for disconnecting their freezers for a few days or more to investigate the matter

Kind Regards John

 

[ericmark]

It would seem they do exist and we had some thing similar fitted to our batching plants. It the plant tripped out full it would have been a mammoth job to have empted it before restart so there was just a monitor so it would not be re-started before fault was corrected.

Although the write up on X-Pole RCD has be up for some time trying to find an outlet does not seem that easy Farnell do seem to do them. As do RS Components but the crunch comes with the £93.60 price tag.

This is likely the problem is we are not prepared to spend out on the devices which monitor in fact before 17th Edition forced our hand few domestic properties had RCD protection on all circuits.

What is wanted is something looking like one of the plug in energy meters so we can return and monitor latter and if set to flag any leakage over 3.5mA that would be enough as anything over 3.5mA should not be using a standard 13A plug anyway.

My PAT tester has a 0 - 6 mA meter and shows green up to 3 mA and is the only way I can test leakage at the moment the clamp-on scale will not read mA.

The clamp-on earth leakage tester Milliamp clamp on able to measure the milliamp are not cheap at around £200 but has a Data hold memory function so would do the job. So it's all down to how much one is willing to pay.

For a large firm one between 10 electricians would seem a good idea but for the sole trader rather expensive considering how often used.

 

[bernardgreen]

The sure way to determine what was causing an RCD to trip would be to monitor and record the difference between Live and Neutral on every circuit with a resolution of 10 milli seconds or better.

A current transformer of every cable leaving the CU would almost certainly detect where the leakages were the next time the RCD tripped.

Fortunately the amount of data to be retained for analysis is only the two or three seconds before the trip so a massive data store is not needed. When the RCD trips the data storage freezes.

To do all circuits at the same time would be difficult due to findind space for all the transformers so maybe a couple of cables at a time and if the data for the first trip did not show anything then move the transformers to another set of cables.

 

[JohnW2]

The sure way to determine what was causing an RCD to trip would be to monitor and record the difference between Live and Neutral on every circuit with a resolution of 10 milli seconds or better.

Well, yes, but for the sort of domestic troubleshooting we're talking about, to simultaneously monitor every circuit (and presumably every individual load on every circuit!) would be totally impractical/unrealistic!

Kind Regards, John

 

[ericmark]

I can find 10 mA RCD's

but no price. I can even find a PDF with all details but not the device for sale on UK priced site.

 

[JohnW2]

What is wanted is something looking like one of the plug in energy meters so we can return and monitor latter and if set to flag any leakage over 3.5mA that would be enough as anything over 3.5mA should not be using a standard 13A plug anyway.

That's essentially what I am suggesting/proposing - and if the market became large enough, I don't see why it would need to be particularly expensive. It could probably be produced using little more than the components present in a cheap electronically-triggered RCD, or very similar components.

Kind Regards, John

 

[EFLImpudence]

I didn't even know that they existed - but, yes, it sounds as if they could probably be used (or adapted to be used) for the purpose. What exactly are they - just a low current clamp meter for a single conductor, or do they (like an RCD) clamp over an L/N pair and detect an imbalance? If not the later, do I take it that you'd use it on the CPC (which you would obviously have to 'expose' somehow). That obviously would not work if it were, say, a washing machine which was leaking current to earth primarily through pipework.

http://www.fluke.com/Fluke/usen/Clamp-Meters/Fluke-360.htm?PID=56072&trck=360

Yes, they are just a high (or is it low?) resolution residual current clamp meter (down to 1mA or even microA but then really expensive) which should be clamped around L & N.

As you say, around CPC is not reliable.

You can make a short extension lead with the cores exposed.

 

[JohnW2]

I can find 10 mA RCD's

but no price. I can even find a PDF with all details but not the device for sale on UK priced site.

Well, Here is an in-line' one for £14.26 + VAT. You could easily add a plug and trailing socket for a pound or two more.

Kind Regards, John

 

[plugwash]

I can find 10 mA RCD's

but no price. I can even find a PDF with all details but not the device for sale on UK priced site.

Here is one on a UK site though it's inline rather than plugtop

http://cpc.farnell.com/pro-elec/inlinercd/rcd-in-line/dp/PL10474?Ntt=10ma+rcd

 

[JohnW2]

http://www.fluke.com/Fluke/usen/Clamp-Meters/Fluke-360.htm?PID=56072&trck=360
Yes, they are just a high (or is it low?) resolution ...

I suspect you probably don't actually mean either It may or may not be 'high resolution', but I suspect what you actually mean is 'high sensitivity'

... residual current clamp meter (down to 1mA or even microA but then really expensive) which should be clamped around L & N. As you say, around CPC is not reliable.

Fair enough. Thanks.

You can make a short extension lead with the cores exposed.

Indeed one can. I have several variants thereof in my toolbox

Kind Regards, John

 

[JohnW2]

Here is one on a UK site though it's inline rather than plugtop
http://cpc.farnell.com/pro-elec/inlinercd/rcd-in-line/dp/PL10474?Ntt=10ma+rcd[/QUOTE]
Snap!

Kind Regards, John