I.R. testing existing circuits

[LastManOnline]

The accepted method of I.R. testing existing circuits has been to join L and N together and test to E, to protect against damaging any electronics that were inadvertently left in circuit.
However ,with the increase of new types of circuits between L and E as a result of "intentional leakage" built in to appliances , will we need to review how we carry out I.R.testing in the near future. ?

 

[Dustydazzler]

For most part and this is just my own opinion is that and IR test should be done for initial verification ( commissioning the install )
Do we really need to IR test an install as part of any further EICRs ? Only I guess if something looks dodgy or you have some sort of intermittent fault for example.
I honestly don't feel IR tests are needed as part of any sort of basic periodic inspection

 

[DPG]

For most part and this is just my own opinion is that and IR test should be done for initial verification ( commissioning the install )
Do we really need to IR test an install as part of any further EICRs ? Only I guess if something looks dodgy or you have some sort of intermittent fault for example.
I honestly don't feel IR tests are needed as part of any sort of basic periodic inspection

But surely if the IR of a circuit is reducing at each inspection then that is a sure sign of a problem such as a cable break down. An IR test is a key test of the condition of an installation.

 

[Pretty Mouth]

The accepted method of I.R. testing existing circuits has been to join L and N together and test to E, to protect against damaging any electronics that were inadvertently left in circuit.
However ,with the increase of new types of circuits between L and E as a result of "intentional leakage" built in to appliances , will we need to review how we carry out I.R.testing in the near future. ?

Are you coming across many otherwise healthy appliances giving very low IR because of this? I can't say I've come across any myself, that I've noticed. Sometimes lowish IR, perhaps bringing the circuit down to 1 Mohms, but never low enough that I'm especially concerned by it.

 

[Lucien Nunes]

the increase of new types of circuits between L and E as a result of "intentional leakage" built in to appliances ,

Most functional leakage is due to capacitance e.g. from filter capacitors connected L-E, and from parts of a circuit that operate at much higher frequencies than 50Hz. Both of these will only occur when the equipment is connected to AC, not when subjected to a DC insulation test.

As an example, many PAT results on larger electronic appliances will show maybe 0.5-1mA of differential leakage. A 1mA resistive leakage would indicate a resistance of 230/0.001 = 0.23MΩ. Yet an IR test on the same device will usually show tens or hundreds of megohms, proving that the 1mA is capacitive in origin. The lowest actual resistance between L & E that I have seen incorporated into standard electronic goods, that would be revealed by an IR test, was 7MΩ in certain models of Sony professional video monitor.

 

[pc1966]

The lowest actual resistance between L & E that I have seen incorporated into standard electronic goods, that would be revealed by an IR test, was 7MΩ in certain models of Sony professional video monitor.

I found some RCD sockets at my parents were each around 50k L+N to E, almost certainly a deliberate thing to force-trip if N is live/open.

I initially panicked though as seeing around 0.02M (but not a dead short) when doing a quick check made me think the 50+ year old cable had gone very, very bad!

 

[LastManOnline]

Are you coming across many otherwise healthy appliances giving very low IR because of this?

No.Like you I get the expected lowish IR readings associated with all appliances.But what I am really curious about is whether any one has experienced issues where electronics may have been damaged while carring out I.R. testing between L and E ?( I have,nt )

 

[IAmSparkytus!]

Are you coming across many otherwise healthy appliances giving very low IR because of this? I can't say I've come across any myself, that I've noticed. Sometimes lowish IR, perhaps bringing the circuit down to 1 Mohms, but never low enough that I'm especially concerned by it.

Just out of interest would you not be concerned by a reading as low as 1 or even 2 megohm?

We only really carry out new install so more often than not our readings are always as high as our testers can go which is >999

Whenever I see a reading which is below 100 even I instinctively start to wonder how long before there’s a big problem!

 

[Lucien Nunes]

experienced issues where electronics may have been damaged while carring out I.R. testing between L and E ?( I have,nt )

No, and it's unlikely. It would only really occur where the test is carried out from one live conductor to earth, the other one is shorted or very leaky to earth, and the load is very small and very sensitive. Nonetheless, if one uses a method that never puts 500V across the load, one can never be accused (rightly or falsely) of damaging it as a result.

Whenever I see a reading which is below 100 even I instinctively start to wonder how long before there’s a big problem!

On ten feet of new 2.5 T+E sat on a warm dry bench, I would want to know why it only achieved such a low figure. On a typical installation in service, not so much.

 

[Dave OCD]

We did a follow up on an EICR including a CU change where a landlord 'didn't like the attitude' of the previous Electrician. It was an unsatisfactory report and the only RFC for the house had a tested IR of 0.28 Megohms across L-E which he'd given a 'FI' code to. There was also a C2 for one socket with reverse polarity - having identified this how long would it take to rectify there and then ffs ?!
We had a good look around and found a little concrete shed in the back garden with a twin RCD socket spurred off the ring, very neatly done actually. Removed this and the ring IR was over 100 Megohms on all conductors.

 

[pc1966]

Whenever I see a reading which is below 100 even I instinctively start to wonder how long before there’s a big problem!

I don't think I would worry at 100M, maybe at 10M if a small and clean looking install.

But the usual limits of 0.5/1M do seem worryingly low to me unless I could see some very reasonable explanation, maybe like a big hall of many damp and dusty lights, etc.

 

[Pretty Mouth]

No.Like you I get the expected lowish IR readings associated with all appliances.But what I am really curious about is whether any one has experienced issues where electronics may have been damaged while carring out I.R. testing between L and E ?( I have,nt )

I haven't either, however, a L-E test at 500V could be a problem if there was a N to E fault present. With N and E shorting, you would inadvertently be putting 500V across L and N. Same if there's a borrowed neutral, and the circuit borrowed from was not isolated.

Just out of interest would you not be concerned by a reading as low as 1 or even 2 megohm?

We only really carry out new install so more often than not our readings are always as high as our testers can go which is >999

Whenever I see a reading which is below 100 even I instinctively start to wonder how long before there’s a big problem!

My work's mainly minor works and fault finding on existing domestic installations, where low IR is normal, and mostly from plugged in appliances etc. Actual faults are usually right down at <0.1Mohm. I don't usually bother unplugging anything these days, unless the circuit is less than 1Mohm.

OTOH, if I was getting 1 or 2 Mohms on a circuit that I'd just installed, then yes, I'd be looking for the cause.

 

[cliffed]

IR testing is always an issue on existing installations but necessary
In large premises maybe the option would be an Earth Leakage meter on the supply to measure the IR

 

[littlespark]

I did a fault find last week.

Customer panicked as she had no socket power and a work Zoom call. In 10 minutes, I found the faulty circuit, disconnected it and got the remaining circuits live. (6 socket circuits on one rcd…. The type you need to switch to reset before it can be turned on again)

Went back Saturday morning… fault in conservatory. 20A radial. 5 double sockets.
Thought it was the one next to the dog cage, as the floor was soaking wet where they tipped their bowl. Split circuit and IR test. Clear one way, very low other way.

Tried at the other sockets and of course, the last one I opened, the one behind all the heavy furniture was the one at fault… and happened to be the first on the circuit from the board…

The cause of the fault was a rather large slug resting itself across the terminals of the socket…
The back of the socket itself was shining in slug slime.

 

[LastManOnline]

I haven't either, however, a L-E test at 500V could be a problem if there was a N to E fault present. With N and E shorting, you would inadvertently be putting 500V across L and N. Same if there's a borrowed neutral, and the circuit borrowed from was not isolated.

A point well worth emphasising this.The sequencing of I.R. tests on existing installs.Always testing N to E first.
Of course many of these faults should be picked up by rcd,s , but not all circuits are rcd protected and even when they are ,the leakage could be under the rcd threshold.
I also strongly recommend double checking that tester is on 250 volt first. 500 volt later.

 

[Dustydazzler]

No, and it's unlikely. It would only really occur where the test is carried out from one live conductor to earth, the other one is shorted or very leaky to earth, and the load is very small and very sensitive. Nonetheless, if one uses a method that never puts 500V across the load, one can never be accused (rightly or falsely) of damaging it as a result.


On ten feet of new 2.5 T+E sat on a warm dry bench, I would want to know why it only achieved such a low figure. On a typical installation in service, not so much.

Reminds me of my college days when ever we did IR tests on our test circuits we always got 999MO because everything was wired in such short cable. The teacher kept having to remind us in the 'Real World' we should not expect such a high IR reaching.

Out of curiosity what would the average Sparks deem an unacceptable IR reading for an in service install , 2MO 3MO would you want to investigate further ?

 

[LastManOnline]

Out of curiosity what would the average Sparks deem an unacceptable IR reading for an in service install , 2MO 3MO would you want to investigate further ?

While acceptable from the regs point of view ,in practice you would be unhappy, possibly concerned with such readings.There is certainly trouble on the way.

 

[littlespark]

Reminds me of my college days when ever we did IR tests on our test circuits we always got 999MO

My college days we did IR tests on other apprentices!

 

[Dustydazzler]

My college days we did IR tests on other apprentices!

Not at college but one of my first weeks on-site doing a big PIR on an old house the lead sparks got me to drop a socket off and told me to hold onto the bare cpcs and he went off and did an IR test at the board. I should have known better but when you are green you do what you are told. I wonder what reading he got with me in the circuit

 

[westward10]

If you hold the ends and press the button once the initial jolt happens you can't really feel it even on 1kv.

 

[littlespark]

We used an old wind up tester at college… it was a test to see who could hold the lead pins the longest as it was wound faster and faster.

Fingertips went numb, and the pins just slipped out.

 

[LastManOnline]

For most part and this is just my own opinion is that and IR test should be done for initial verification ( commissioning the install )
Do we really need to IR test an install as part of any further EICRs ? Only I guess if something looks dodgy or you have some sort of intermittent fault for example.
I honestly don't feel IR tests are needed as part of any sort of basic periodic inspection

I,m slightly off topic here ,but I felt I should say that I,m largely in agreement with what you say.Certainly ,I cant see the point of extensive electrical testing of a 5 year old family home ,when in reality a thorough visual inspection should suffice (plus testing of all safety devices).
Now a 50 year old home ?Thats a very different matter.
But whether 5 or 50 ,I regard extensive visual checks as my number one source of information in determining the health of an electrical installation.Standard electrical testing then supplements it.
Did a periodic on Wednesday on a recently renovated older house which is up for sale.Looked great and I felt optimist.All the standard testing gave excellent results.
When I began removing accessories though I discovered undersized wiring at many sockets.Secondly the new wiring visible at the DB often extended only to the first point on the circuit where it was joined to the original wiring.It all looked professionally done too.It set off alarm bells immediately.
I dont believe it possible to experience "a breakdown of insulation" in a modern home,certainly not through deterioration.And as Dusty alludes to the insulation is only really compromised (in my experience to) when some unconventional activity has occured in the home ,be it intentional or unintentional and this will almost always manifest itself either visually ,audibly ,via odour or through operation of one of the safety devices.

 

[Simon47]

Most functional leakage is due to capacitance e.g. from filter capacitors connected L-E, and from parts of a circuit that operate at much higher frequencies than 50Hz. Both of these will only occur when the equipment is connected to AC, not when subjected to a DC insulation test.

The RCBOs we have in our rental properties have about 65k to earth one polarity, >999M the other - so can only do in-circuit tests at one DC polarity. But overall there's a fair bit of capacitance in the installation such that my MFT can't get a decent reading. So I break out dad's old "press and hold - wait till the needle stabilises" meter, and with that I can do the whole installation at once which saves a lot of work dropping tails out of the breakers (and the associated risks of not getting them back in properly).

No.Like you I get the expected lowish IR readings associated with all appliances.But what I am really curious about is whether any one has experienced issues where electronics may have been damaged while carring out I.R. testing between L and E ?( I have,nt )

I've never seen anything, and up to 350V DC (peak voltage of 250V sine wave) it's hard to see how something survived on mains power anyway. Between 350V and 500V really isn't that much, so I'd have thought anything so marginal would get knocked out by the occasional surge anyway. So personally I'd not have any concerns testing at 500V unless I knew there was something both sensitive and expensive.

Whenever I see a reading which is below 100 even I instinctively start to wonder how long before there’s a big problem!

I know what you mean - it just "feels wrong" even though we know it's OK according to the regs. Even down to 1M, that's only 0.25mA and a grand total of ... drumroll ... 62mW of power being dissipated along "a fair bit" of cable. But as you say, for something that at one time would have been too high for our meters to measure, why is is so much worse now ?