Best way to confirm correct circuit connections?
Best way to confirm correct circuit connections?
Just curious what methods people use to confirm this
Re: Best way to confirm correct circuit connections?
For new work, the basic method I use is to connect earth conductors at the switchboard first, then do the earth continuity test with a trailing lead.
That confirms that the earth is connected correctly. (and completes the earth continuity test).
This is done after all socket outlets, lights etc are fitted off of course. If other items like oven, cooktop, hot water aren't fitted off at that stage, they can be tested during their fit off by disconnecting the circuit neutral (which is necessary for the IR test anyway).
After the earth continuity test, I would do the IR test between A & N (shorted together) to earth before connecting each circuit at the switchboard. Testing for interconnections can be done at the switchboard at this stage as well.
After the earth conductor is confirmed, the A & N polarity can be tested with a normal voltage test after livening. Some MFTs have a button to switch the voltage test between A-E, N-E, A-N.
Confirming that active is switched can be done during the voltage test, or with a volt stick.
You may have noticed that polarity and correct circuit connections tests have been combined in section 8 of 3000:2018 which was sensible. As in my description above, some testing can cover more than one aspect of the overall requirements.
That confirms that the earth is connected correctly. (and completes the earth continuity test).
This is done after all socket outlets, lights etc are fitted off of course. If other items like oven, cooktop, hot water aren't fitted off at that stage, they can be tested during their fit off by disconnecting the circuit neutral (which is necessary for the IR test anyway).
After the earth continuity test, I would do the IR test between A & N (shorted together) to earth before connecting each circuit at the switchboard. Testing for interconnections can be done at the switchboard at this stage as well.
After the earth conductor is confirmed, the A & N polarity can be tested with a normal voltage test after livening. Some MFTs have a button to switch the voltage test between A-E, N-E, A-N.
Confirming that active is switched can be done during the voltage test, or with a volt stick.
You may have noticed that polarity and correct circuit connections tests have been combined in section 8 of 3000:2018 which was sensible. As in my description above, some testing can cover more than one aspect of the overall requirements.
Re: Best way to confirm correct circuit connections?
I pretty much do the exact same as above, I e just never really understood correct circuit connections
Confirm EC with N disconnected
Proves E in right place
IR test
A to E then N to E just because I just find this faster than linking them
I also do A to N if possible, at 500V if certain nothing connected or at 250V if uncertain usually back to 500V once 250V comes back clear
Connect N and I'll then test continuity between my known E I confirmed in my EC test and N with all switches off, that confirms my N polarity and that no switches in N
Liven
Confirm voltage on A to E and check switches remove voltage
Check EFLI, I usually do this regardless if I can although only mandatory on non-RCD S/Os just to back up my EC value and confirm good mains PEN
Check RCDs if installed, test N to load A 230V, N to load N continuity, hit test and check same but no voltage and no continuit
Confirm EC with N disconnected
Proves E in right place
IR test
A to E then N to E just because I just find this faster than linking them
I also do A to N if possible, at 500V if certain nothing connected or at 250V if uncertain usually back to 500V once 250V comes back clear
Connect N and I'll then test continuity between my known E I confirmed in my EC test and N with all switches off, that confirms my N polarity and that no switches in N
Liven
Confirm voltage on A to E and check switches remove voltage
Check EFLI, I usually do this regardless if I can although only mandatory on non-RCD S/Os just to back up my EC value and confirm good mains PEN
Check RCDs if installed, test N to load A 230V, N to load N continuity, hit test and check same but no voltage and no continuit
Re: Best way to confirm correct circuit connections?
I've just had a look at 2018 and I think I'm actually pretty good
I think I was just always confused by polarity vs correct circuit connections but seeing them combined and reading through the lastest version it seems a bit more clearer of what's needed to be achieved and I think I pretty much cover that
I think I was just always confused by polarity vs correct circuit connections but seeing them combined and reading through the lastest version it seems a bit more clearer of what's needed to be achieved and I think I pretty much cover that
Re: Best way to confirm correct circuit connections?
Yes there is some logic to treating "correct circuit connections" as being part of "polarity".
But there are aspects of ccc that have nothing whatever to do with polarity.
So the change (in A1 to 2018, not in original 2018) makes some aspects more logical; but risks causing other confusions.
As Doug suggests; the various types of testing required by Section 8 are not a set of unrelated tests.
They are all inter-dependent; and when separating them into groups / headings, they can be grouped in many different ways.
Yes there are close ties between some aspects of ccc and some aspects of polarity; which is why the Committee amalgamated these two headings.
But there are also ties with other types / headings of testing.
(referring now to 2007 +A1 +A2)
8.3.8.1 declares the purposes of ccc testing to be
a) that protective earthing conductors do not normally carry load current; and
b) avoiding damage / injury that would arise from the high currents that result from a short circuit.
While it's true that incorrect polarity can result in PECs carrying load current, so can other errors.
- eg; having a 2nd N-E connection, which used to be common practice but is no longer allowed
(and this avoidance of load current flowing on PECs is one that there is no specific rule about other than in Section 8)
Similarly for short circuits, they are not all caused by incorrect polarity.
Looking at the required results;
a) a short can exist between conductors of a circuit even when polarity is correct.
b) sometimes transposition of conductors can result in incorrect polarity; but other transpositions cause incorrect phase rotation.
- both these conditions do relate to polarity, specifically to 8.3.7.1 (a), (b), & (d).
However c) - interconnection between conductors of different circuits - has nothing at all to do with polarity.
And nothing mentioned in 8.3.8 ties back to correct polarity of switching.
The key is to look at all the stated purposed, and all the required outcomes, across all the subheadings of 8.3.
A testing regime, taken over all, has to show that nothing has gone wrong WRT any of these aspects.
This is usually thought of as (as per 8.3.1) verifying that the work done complies with requirements.
As a result, many people consider what (for example) 'correct polarity' looks like, and base their testing on showing that 'correct polarity' has been achieved.
This is a flawed approach ; particularly for polarity testing.
Instead what we need to do is consider all the many different ways in which polarity might have become incorrect; and apply testing that shows that none of those faults has occurred.
This is one of the factors behind the recently publicised cases of reverse polarity of mains.
In both cases highlighted by Worksafe; , and as is common throughout our industry;
the methods adopted were fundamentally flawed - in at least two ways.
Firstly they failed to take proper precautions for safety during the testing; in that they assumed there would be nothing wrong and took no measures to mitigate the consequences if something was wrong.
Secondly the methods (voltage to independent earth) were based on looking for what 'correct polarity' looks like, and failing to recognise that reverse polarity can actually look very similar.
The specific tests necessary depend in part on what we are testing.And the methodology can be adapted according to circumstance.
Testing a new final subcircuit is different from testing an entire new installation.
That's why we are not required to follow the specific tests specified in "3017";
instead we can vary them - but only provided the results are equally valid.
The problem is that most practitioners are simply not competent to assess the validity of alternative test methods;
so any variation should be restricted to things that can't make a difference.
Even an apparently minor issue such as IR testing [A+N - E] or [A-E] then [N-E] can have consequences.
Usually won't cause any harm, or any incorrect result ; but under some fault conditions it can do either / both.
-----------------
Both the test sequences suggested above appear designed for a single new subcircuit - but they do not actually test all aspects as is required.
For example, there's no testing for interconnections between different circuits.
Jamie's test of continuity E - N with all switches off may show "no switches in N"; however this is not actually the requirement.
What it certainly does not do is show correct polarity of the N ; ie that the N is connected to the correct terminal(s) of equipment.
It is a partial test only, and can only show no N-E transposition on the circuit tested.
There could be an A-N reversal, but this test won't find it.
Doug's option of testing voltage are socket / appliance after livening is also flawed, as it won't detect N-E transposition.
Or rather, not by itself. There's NO live test that can do that as long as the MEN is in place.
However in combination with the earlier ECC test, and - as he suggests - test of switch operation removing power; we get (single phase) polarity
However consider possible combinations of faults.
What if there was a A-N transposition somewhere along the circuit, and a switch incorrectly connected into the N (ie not also switching A).
You'd get a set of voltage readings exactly what "correct" would look like; and the switch would remove the power.
Jamies' RCD test correctly tries to show - as required - that operation of the RCD has disconnected the circuit.
However it's only a partial rest.
First you need to have shown that the RCD has been correctly connected, especially WRT line / load and A / N.
Not all are sensitive to both, but most have a dedicate N pole that has to be correct and some have defined line & load terminals.
There's risk of the conductors on the A & N load terminals being from different circuits; crossed Ns being surpisingly common.
So that needs to be proven before continuing.
Then you have to cause the operation (the only bit most people bother with).
Then, to show proper disconnection, you need to do 4 tests (for a 2P single-phase RCD).
1/. test A-N line (still 230 V);
which shows that the vanishing of the voltage readout was due to operation of the RCD, and not to some other co-incidental cause).
2/. test A-N load; 0 V means the either the A or the N pole has opened , maybe both.
3/. test line N to load A; 0V shows A pole open.
4/. test line A to load N; 0V shows N pole open.
Like other test methods in "3017" , this one (new in 2022 edition) only gives valid results provided the assumptions it is based on are confirmed to be correct. Hence the need to prove line / load correct, A / N correct, and that all conductors are from same circuit.
Remember it's not about testing the RCD itself; it's about showing that it has been correctly connected into the circuit.
Once again: what might have gone wrong, and how can we show that none of those things has actually gone wrong - whether singly or in combination with other faults.
Yes, the opening of the poles of the RCD can be done using continuity instead of voltage.
But because the middle step (causing operation) requires that the RCD be energised,and leaves the line terminals energised; best to avoid putting an ohm meter into the equation.
--------------------------
The methods set out in "3017" may take a bit longer than some of the shortcuts we may be used to; but they have been thought through so that, in combination, they identify any faults.
And the new edition - out soon - should be easier to follow.
But there are aspects of ccc that have nothing whatever to do with polarity.
So the change (in A1 to 2018, not in original 2018) makes some aspects more logical; but risks causing other confusions.
As Doug suggests; the various types of testing required by Section 8 are not a set of unrelated tests.
They are all inter-dependent; and when separating them into groups / headings, they can be grouped in many different ways.
Yes there are close ties between some aspects of ccc and some aspects of polarity; which is why the Committee amalgamated these two headings.
But there are also ties with other types / headings of testing.
(referring now to 2007 +A1 +A2)
8.3.8.1 declares the purposes of ccc testing to be
a) that protective earthing conductors do not normally carry load current; and
b) avoiding damage / injury that would arise from the high currents that result from a short circuit.
While it's true that incorrect polarity can result in PECs carrying load current, so can other errors.
- eg; having a 2nd N-E connection, which used to be common practice but is no longer allowed
(and this avoidance of load current flowing on PECs is one that there is no specific rule about other than in Section 8)
Similarly for short circuits, they are not all caused by incorrect polarity.
Looking at the required results;
a) a short can exist between conductors of a circuit even when polarity is correct.
b) sometimes transposition of conductors can result in incorrect polarity; but other transpositions cause incorrect phase rotation.
- both these conditions do relate to polarity, specifically to 8.3.7.1 (a), (b), & (d).
However c) - interconnection between conductors of different circuits - has nothing at all to do with polarity.
And nothing mentioned in 8.3.8 ties back to correct polarity of switching.
The key is to look at all the stated purposed, and all the required outcomes, across all the subheadings of 8.3.
A testing regime, taken over all, has to show that nothing has gone wrong WRT any of these aspects.
This is usually thought of as (as per 8.3.1) verifying that the work done complies with requirements.
As a result, many people consider what (for example) 'correct polarity' looks like, and base their testing on showing that 'correct polarity' has been achieved.
This is a flawed approach ; particularly for polarity testing.
Instead what we need to do is consider all the many different ways in which polarity might have become incorrect; and apply testing that shows that none of those faults has occurred.
This is one of the factors behind the recently publicised cases of reverse polarity of mains.
In both cases highlighted by Worksafe; , and as is common throughout our industry;
the methods adopted were fundamentally flawed - in at least two ways.
Firstly they failed to take proper precautions for safety during the testing; in that they assumed there would be nothing wrong and took no measures to mitigate the consequences if something was wrong.
Secondly the methods (voltage to independent earth) were based on looking for what 'correct polarity' looks like, and failing to recognise that reverse polarity can actually look very similar.
The specific tests necessary depend in part on what we are testing.And the methodology can be adapted according to circumstance.
Testing a new final subcircuit is different from testing an entire new installation.
That's why we are not required to follow the specific tests specified in "3017";
instead we can vary them - but only provided the results are equally valid.
The problem is that most practitioners are simply not competent to assess the validity of alternative test methods;
so any variation should be restricted to things that can't make a difference.
Even an apparently minor issue such as IR testing [A+N - E] or [A-E] then [N-E] can have consequences.
Usually won't cause any harm, or any incorrect result ; but under some fault conditions it can do either / both.
-----------------
Both the test sequences suggested above appear designed for a single new subcircuit - but they do not actually test all aspects as is required.
For example, there's no testing for interconnections between different circuits.
Jamie's test of continuity E - N with all switches off may show "no switches in N"; however this is not actually the requirement.
What it certainly does not do is show correct polarity of the N ; ie that the N is connected to the correct terminal(s) of equipment.
It is a partial test only, and can only show no N-E transposition on the circuit tested.
There could be an A-N reversal, but this test won't find it.
Doug's option of testing voltage are socket / appliance after livening is also flawed, as it won't detect N-E transposition.
Or rather, not by itself. There's NO live test that can do that as long as the MEN is in place.
However in combination with the earlier ECC test, and - as he suggests - test of switch operation removing power; we get (single phase) polarity
However consider possible combinations of faults.
What if there was a A-N transposition somewhere along the circuit, and a switch incorrectly connected into the N (ie not also switching A).
You'd get a set of voltage readings exactly what "correct" would look like; and the switch would remove the power.
Jamies' RCD test correctly tries to show - as required - that operation of the RCD has disconnected the circuit.
However it's only a partial rest.
First you need to have shown that the RCD has been correctly connected, especially WRT line / load and A / N.
Not all are sensitive to both, but most have a dedicate N pole that has to be correct and some have defined line & load terminals.
There's risk of the conductors on the A & N load terminals being from different circuits; crossed Ns being surpisingly common.
So that needs to be proven before continuing.
Then you have to cause the operation (the only bit most people bother with).
Then, to show proper disconnection, you need to do 4 tests (for a 2P single-phase RCD).
1/. test A-N line (still 230 V);
which shows that the vanishing of the voltage readout was due to operation of the RCD, and not to some other co-incidental cause).
2/. test A-N load; 0 V means the either the A or the N pole has opened , maybe both.
3/. test line N to load A; 0V shows A pole open.
4/. test line A to load N; 0V shows N pole open.
Like other test methods in "3017" , this one (new in 2022 edition) only gives valid results provided the assumptions it is based on are confirmed to be correct. Hence the need to prove line / load correct, A / N correct, and that all conductors are from same circuit.
Remember it's not about testing the RCD itself; it's about showing that it has been correctly connected into the circuit.
Once again: what might have gone wrong, and how can we show that none of those things has actually gone wrong - whether singly or in combination with other faults.
Yes, the opening of the poles of the RCD can be done using continuity instead of voltage.
But because the middle step (causing operation) requires that the RCD be energised,and leaves the line terminals energised; best to avoid putting an ohm meter into the equation.
--------------------------
The methods set out in "3017" may take a bit longer than some of the shortcuts we may be used to; but they have been thought through so that, in combination, they identify any faults.
And the new edition - out soon - should be easier to follow.
- Rating: 33.33%
Re: Best way to confirm correct circuit connections?
Understood and noted. Very much look forward to reading the new version and improving how I can.
However, just curious what I'm missing in regards to above, because I test EC with a trailing lead and I have my N disconnected this to be should confirm EC value but also E polarity is correct because I know E is in the right place then wouldn't a N-E continuity test (with N connected again) confirm my N is in the right place and not transposed with A or an A. Then simply checking voltage from A to either once power is on would confirm connection of the A?
However, just curious what I'm missing in regards to above, because I test EC with a trailing lead and I have my N disconnected this to be should confirm EC value but also E polarity is correct because I know E is in the right place then wouldn't a N-E continuity test (with N connected again) confirm my N is in the right place and not transposed with A or an A. Then simply checking voltage from A to either once power is on would confirm connection of the A?
Re: Best way to confirm correct circuit connections?
yes, I missed that you have the N disconnected for the ECC test.
Which is the equivalent of having it connected, but MEN open.
In that case, yes the ECC testing shows PECs correctly connected at all points along circuit
(as well as confirming the PEC part of the EFLI (which for many circuits is enough confirmation of EFL; no full test needed) .
Then, once the N & E are connected at source end; testing N-E loop continuity can confirm N correctly connected at all points.
Having the N-E loop open is a key part of polarity testing;
Which is the equivalent of having it connected, but MEN open.
In that case, yes the ECC testing shows PECs correctly connected at all points along circuit
(as well as confirming the PEC part of the EFLI (which for many circuits is enough confirmation of EFL; no full test needed) .
Then, once the N & E are connected at source end; testing N-E loop continuity can confirm N correctly connected at all points.
Having the N-E loop open is a key part of polarity testing;
Re: Best way to confirm correct circuit connections?
While I totally agree, I would be very careful with wording like the above.
So many people I come across in discussion groups say that people should "remove the MEN link when testing".
It's definitely something that I have to keep correcting people on.
There's really only two test methods (of the mandatory tests) that come to mind that require the MEN link to be removed.
An IR test for a whole installation.
Initial live polarity test for mains.
In all other cases where the live conductors of a circuit need to be tested (active or neutral), they should be isolated, and the circuit neutral disconnected if necessary to carry out the tests.
As an example I use for people - if you were testing a few new circuits or a submain in a large commercial installation, you certainly wouldn't be removing a MEN link for the whole installation just to test those circuits.
And also point out that removing the MEN link on a live installation (or livening with the MEN link out) should only be done if the proper precautions are taken.
Re: Best way to confirm correct circuit connections?
I (carefully) didn't say "having the MEN open is key"; I said having the N-E loop open is key.
Which may be by opening the MEN, or by disconnecting the circuit N (but generally should not be achieved by disconnecting any PECs).
Unless we take this step, we can't eliminate possibility of N-E reversal (when downstream of the MEN).
And for polarity of PEN wiring systems (whether mains or submains) we also need to open the N-E loop.
I agree opening the MEN to test circuit(s) added to an existing installation would probably not be appropriate; for a number of reasons.
Would generally be better to connect E, and leave A &N 'disconnected' while testing IR & polarity / ccc.
Remember we also need to test for interconnections between our new circuit and the rest of the installation.
This mandatory test is often skipped, just because we 'know" we haven't created any.
Doing doesn't take much time / effort; but it requires our new circuit A(s) & N to not be connected to the rest of the installation (which could be by switching off the RCBO).
Also agree about care needed when opening an MEN; even on a fully "isolated" installation.
When livening installations the MEN should (always?) be open, as well as all main switches being off;
to avoid the installation earthing system being energised by incorrect polarity of supply.
Alternatively the incoming mains N can be disconnected, but this carries a whole other set of risks.
This is exactly what went wrong on the two cases recently highlighted by Worksafe.
Important that the methods we use are not only adequate to confirm no faults;
but also must be such as to minimise risks to both those doing the testing, and others in the vicinity .
Clause 8.1.2; para 3. Also ESR13 (3).
Which may be by opening the MEN, or by disconnecting the circuit N (but generally should not be achieved by disconnecting any PECs).
Unless we take this step, we can't eliminate possibility of N-E reversal (when downstream of the MEN).
And for polarity of PEN wiring systems (whether mains or submains) we also need to open the N-E loop.
I agree opening the MEN to test circuit(s) added to an existing installation would probably not be appropriate; for a number of reasons.
Would generally be better to connect E, and leave A &N 'disconnected' while testing IR & polarity / ccc.
Remember we also need to test for interconnections between our new circuit and the rest of the installation.
This mandatory test is often skipped, just because we 'know" we haven't created any.
Doing doesn't take much time / effort; but it requires our new circuit A(s) & N to not be connected to the rest of the installation (which could be by switching off the RCBO).
Also agree about care needed when opening an MEN; even on a fully "isolated" installation.
When livening installations the MEN should (always?) be open, as well as all main switches being off;
to avoid the installation earthing system being energised by incorrect polarity of supply.
Alternatively the incoming mains N can be disconnected, but this carries a whole other set of risks.
This is exactly what went wrong on the two cases recently highlighted by Worksafe.
Important that the methods we use are not only adequate to confirm no faults;
but also must be such as to minimise risks to both those doing the testing, and others in the vicinity .
Clause 8.1.2; para 3. Also ESR13 (3).
- Rating: 16.67%