Best way to find a short circuit between L and N? I.e a tripping MCB.

[HappyHippyDad]

Finding a fault on a tripping RCD is usually fairly straight forward. The fault usually shows up on an IR test, then we just break the circuit down.

However, with a tripping MCB I find it harder.

We know that a 30mA RCD will trip (approx) with a L/N -E IR of around 8000ohms or below (I'm being very simplistic here, I realise the figure can be much higher as it changes very easily. the point being we have an idea of what to look for) , so we have a figure to keep in mind when testing. Is there a figure we are looking for when an MCB trips, or are we simply looking for continuity?

The reason I find it difficult, is because it can be sometimes impossible to remove all plugged in appliances/FCU's from the offending circuit, meaning we are always going to get a fairly low resistance figure between L and N, sometimes extremely low, low enough to indicate continuity, meaning an IR test for a tripping MCB is not a useful tool for finding the fault (as it is with a tripping RCD). Also, we wouldn't even know if we have removed everything, as there will always be a low resistance figure between L-N because we have a fault.

What are your go to methods for finding the fault linked with a tripping MCB?

I am assuming its not overload as this can be determined by working out max demand. Also, not a L-E fault as this could be found with IR testing. Purely a L-N fault/short circuit.

 

[ProDave]

For a L-N short it in not an insulation tester you need but a low ohms tester, same as you use for instance for measuring continuity on a ring final or doing R1+R2 tests.

To get >32A flowing to trip a 32A mcb requires the "load" (in this case the short circuit) to be less than 7.5 ohms so that should be easy to find.

Same as IR faults start splitting the circuit down into individual legs.

But first ask the question "What work has been done recently" and you mean ANY work done in the house. My favourite example was then the customer said none, I spent the best part of an hour to determine the faulty bit of cable was a light cable above the kitchen ceiling. At that point the customer said "we just had the kitchen flat roof replaced" I rest my case.

 

[Phoenix]

And if it goes instantly with a loud bang, rather than after a number of seconds, you can assume its kicking out on the mag, for a B32 thats going to be at least 3x32A = 96A. So say 100A and do a quick *** packet caluclation, You've got a resistance thats 2.3ohms or below, which is a lot lower than most loads will have (Watch out for things like transformers though, that can have quite a low resistance, because most of their impedance is reactive), but if its as low as 2.3ohms then you'd expect it to sometimes take the B32 MCB out through inrush. So maybe you might see a resistance that low if someone has left a 5ka site transformer or old skool stick welder plugged in, but unlikely in a domestic. So just go round and disconnect the appliances you can see, afterall the issue might be with an appliance. then start trying to work out where on the circuit fault is, if its still there

When you are talking of low resistances like that, the cable resistance is going to be significnant to your readings, so you can use your plug top lead and go round the circuit and find which sockets have the lowest reading and therefore are closest to the fault.

 

[HappyHippyDad]

For a L-N short it in not an insulation tester you need but a low ohms tester, same as you use for instance for measuring continuity on a ring final or doing R1+R2 tests.

To get >32A flowing to trip a 32A mcb requires the "load" (in this case the short circuit) to be less than 7.5 ohms so that should be easy to find.

Same as IR faults start splitting the circuit down into individual legs.

But first ask the question "What work has been done recently" and you mean ANY work done in the house. My favourite example was then the customer said none, I spent the best part of an hour to determine the faulty bit of cable was a light cable above the kitchen ceiling. At that point the customer said "we just had the kitchen flat roof replaced" I rest my case.

And if it goes instantly with a loud bang, rather than after a number of seconds, you can assume its kicking out on the mag, for a B32 thats going to be at least 3x32A = 96A. So say 100A and do a quick *** packet caluclation, You've got a resistance thats 2.3ohms or below, which is a lot lower than most loads will have (Watch out for things like transformers though, that can have quite a low resistance, because most of their impedance is reactive), but if its as low as 2.3ohms then you'd expect it to sometimes take the B32 MCB out through inrush. So maybe you might see a resistance that low if someone has left a 5ka site transformer or old skool stick welder plugged in, but unlikely in a domestic. So just go round and disconnect the appliances you can see, afterall the issue might be with an appliance. then start trying to work out where on the circuit fault is, if its still there

When you are talking of low resistances like that, the cable resistance is going to be significnant to your readings, so you can use your plug top lead and go round the circuit and find which sockets have the lowest reading and therefore are closest to the fault.

Thanks for the in depth answers to both of you.

Yes, I always start with lots of questions for the customer, as it's usually something they have recently done.

So, are you basically saying you're going to get a very low resistance and to use this to split the circuit down and find fault etc? With this way of doing it you just have assume there are no appliances that will have a low enough L-N resistance to send you on a wild goose chase, as there are not many situations where you will have these very low resistance values on appliances?

 

[Fleeting]

So is it a short circuit or not?

 

[HappyHippyDad]

So is it a short circuit or not?

This is just a hypothetical question, for future reference.

 

[ProDave]

And to trip an MCB it is just as likely to be a L-E short, which is much more likely with a nail through a cable.

 

[HappyHippyDad]

And to trip an MCB it is just as likely to be a L-E short, which is much more likely with a nail through a cable.

True, but this feels like it would be easier to track down as you will be testing to earth, so no appliances will be affecting the results.

 

[ProDave]

I can't imagine what "appliances" you are thinking of that cannot be unplugged or switched off at an FCU?

 

[Fleeting]

A short circuit between line and neutral will give a low reading in comparison to an appliance. Using ohms law a 2.3kw (23 ohm) kettle will give 10A so not a short circuit. As already mentioned you would need a low ohm meter and decide if what you are reading could quantify as a dead short, knowing the cable csa you can estimate where it is from the reading.

 

[UNG]

A tone tester that can be set up to operate on open or closed circuit is a good piece of kit using the closed circuit function will only tone to the fault and not beyond it so aiding quickly identifying / locating the fault or shorted location

 

[HappyHippyDad]

I can't imagine what "appliances" you are thinking of that cannot be unplugged or switched off at an FCU?

You must have come across the situation where you are trying to do a L-N IR test, but each time you get a reading, indicating something is still plugged in, or FCU still switched on?
Houses can be huge, with many things hidden over the years. Half the time the customer doesn't even know they've got an old fcu hidden behind a cupboard somewhere, or in the loft feeding a tv aerial amplifier.

 

[Andy™]

I can't imagine what "appliances" you are thinking of that cannot be unplugged or switched off at an FCU?

the ones that are hidden and you don't know they are there

 

[SPECIAL LOCATION]

You must have come across the situation where you are trying to do a L-N IR test, but each time you get a reading, indicating something is still plugged in, or FCU still switched on?

erm... NO....?

As IR tests between all live conductors are normally only required on cables/circuits during new installations, rewires or testing additional wiring to an existing installation.... Where nothing has so far been connected / plugged in!

Whereas, IR testing on existing live installations, (e.g. during Periodic Inspections, or fault investigation), is best done between the live conductors connected together and earth, as per table 3.4, page 129, guidance note3...

Due to the risks of damaging equipment vulnerable to testing...
However tests at 230v can be a useful tool for getting an idea if the circuit has loads of stuff still connected.

So I wouldn't even be using IR tests for investigating short circuit MCB's tripping faults.

Its a meter that can read low resistance continuity in the ohms ranges.. NOT MegaOhms that is your best friend in these situations..

And keep in the back of you mind that a typical 9500watt electric shower has a couple of 12ohm elements in parallel..
i.e. 12ohm in parallel with 12ohm is 6ohm..
AND a lot of shower manufactures still quote their power ratings with ref to 240v Not 230v!! (Check the small print on the box)
So 240v / 6ohm = 40Amps.. 40Amps = 9.6kW at 240v..

SO....
if you are investigating and trying to track down a fault to trip an average 32A MCB socket circuit..
Then you are looking for test readings well below 6ohms..
Therefore IR tests are about as much use as doing RCD trip tests, or checking the external earth loop impedance!!

There was an era before RCD's were required everywhere, and if you were trying to suss out why the 3036 fusewire kept blowing, an insulation resistance tester was not you first choice of meter to use...

e.g. the smallest useful value you can read on the IR test range is probably 0.01MegOhms..
Which is 10,000ohms which is not going to trip ANY MCB's even a 6Amp..

 

[Tim Howard]

As mentioned above a clue can be whether a it’s an instant loud trip (magnetic mechanism) or takes a bit longer (thermal trip)
If you are getting a quick mag trip then there’s a chance of a very very low continuity reading to help track it down.
(Btw if it’s a glorious old Crabtree C50 board they ?uniquely only have mag trips)

After that if you can’t find it it’s down to breaking the circuit down to find out what isn’t causing it.

 

[UNG]

After that if you can’t find it it’s down to breaking the circuit down to find out what isn’t causing it.

And that is where you divide and conquer rather than just start randomly taking the circuit apart
The main faultfinding technique that I was taught as an apprentice that has always stuck is split the circuit in half identify the faulty half and split that in half and keep on repeating until you narrow it down to a cable or joint or termination and then locate the fault

 

[Geoff1946]

A bit of intuitive guesswork can help too. e.g. Are there any outside light fittings? Is there any ongoing building work? Has any area been recently decorated with new accessories?

 

[Andy™]

A bit of intuitive guesswork can help too. e.g. Are there any outside light fittings? Is there any ongoing building work? Has any area been recently decorated with new accessories?

add to that

are there some lights that looks brand new and just installed, complete with all red into one terminal & all black into another, despite never been touched by them and have always worked until yesterday?

 

[binky]

it's also fair to say a L-N short is unlikely, as the CPC is bare copper for a good reason. Ergo, any LN fault is more likely to be an appliance than the cable.

 

[kerching]

Yes there are a lot like that, usually done by someone who has
just emigrated
Got jailed
Head fallen off
Gone offgrid
Run away with a monkey