rcd tripping during the night and insulation resistance testing question

[phil lancaster]

Hi everyone
I have a customer who has a RCD that trips during the night and it resets ok the following day it is ok for a week or so then it tips again during the night.

has any one any ideas and what is the quickest way of finding out what the cause is.
also

I need to do an insulation resistance test an a installation that is totally loaded with equipment that will be damaged if tested and removal is totally impractical any one any advise on a totally safe way to do it
thanks
phil

 

[Andy78]

First thing is to look at things that are most likely triggering the fault in the night. Heating timeclocks, outside PIR lights, off peak supply etc.

 

[ruston]

First thing to check is what is happening through the night that is not happening through the day. Could it be the first activation of a PIR for instance now that the nights are drawing in a bit .
I have not read your profile , so maybe stating the obvious , IR test with LN Joined for vulnerable equipment.

 

[ruston]

Snap Andy.

 

[westward10]

IR test will not damage anything providing you carry out the test correctly.

 

[Wilko]

For no scientific reason, I'm liking external lights on PIR for this. Neighbour's cat runs across the yard... I guess it's TNCS PME and a mix of circuits on that RCD? Gas boiler?

 

[phil lancaster]

Hi I know that insulation resistance testing is normaly carried out with the LN joined but I am always all ways nervous that some damage may sill be done as there is so much sensitive equipment around LEDs Computer eqipiment
phil

 

[Octopus]

Isolate as much as possible and then do proper 3 core ir tests ..... nothing else will work

As the customer to make a note of when it happens ...

I’ve got a similar problem with a customer ... it only trips when they are in, and the kitchen circuit is thought to be the issue .... or an appliance ..... and their DAB radio is the main contender.... it has a plug power adaptor .....

Seen this once before but it was a laptop charger that was to blame ...

 

[ruston]

As a further precaution then test @ 250v

 

[phil lancaster]

And still link L&N and test each circuits at 250 will deff not cause any damage to anything
phil

 

[Deleted member 105166]

Clamp and check the leakage of individual circuits and/or appliances. Also take some min/max leakage readings over a period of time, it could be the combined leakage is often close to the threshold and one appliance used at night is the straw breaking the camel's back.

 

[ruston]

And still link L&N and test each circuits at 250 will deff not cause any damage to anything
phil

Without prejudice lol.

 

[Deleted member 9648]

Times when RCD's trip and incidences of tripping when loads are switched on often mislead and take the inexperienced up the wrong path, especially when faults are N-E. As has already been stated an IR test wont damage anything as long as you keep the test voltage at 250v and do not IR L-N. Bear in mind that if the RCD is tripping this will be a fault/leakage to earth so a L-N test is pointless. Leave all loads in place on the initial test, don't unplug anything. If you get a low reading then start removing loads testing as you go. Lots of people unplug everything before testing, this proves nothing as the cause of the fault may no longer be in place, you want it there when you test, it is a major confidence boost if you quickly establish what you are looking for . 99% of the time RCD tripping can be quickly traced with a logical approach and just an IR tester. I do not possess an earth leakage clamp, while I accept they are useful I trace and rectify faults all the time with just an IR tester.

 

[phil lancaster]

so am I correct in thinking that I should only test the n&e wires and at 250 volts

 

[ruston]

NE joined tested to earth.

 

[westward10]

Test L/N linked at 500v.

 

[Deleted member 9648]

so am I correct in thinking that I should only test the n&e wires and at 250 volts

NO! Test both L-E and N-E...….just avoid L-N.

 

[Deleted member 9648]

so am I correct in thinking that I should only test the n&e wires and at 250 volts

NO! Test both L-E and N-E...….just avoid L-N.

 

[Deleted member 9648]

There's no need to link L-N to IR test to earth for fault finding. The linked L-N method of IR testing is for initial verification of a new/altered circuit, or periodic testing. Fault finding does not follow standard test procedures as required for the above, if you link L-N you will not know whether the fault is L-E or N-E

 

[TJ Anderson]

Test L/N linked at 500v.

If you are just fault finding then I would stick at 250v with loads connected. There is nothing to be gained in this scenario from using 500v IR. You will pick it up on 250v. I know if you link L&N it there is then 0v pd between them but the 250v is a safety net if L or N is opened for any reason in circuit with a termination, switching or MCB etc and you then risk 500v across vulnerable equipment.
I only ever use 500v when I'm 100% all loads are disconnected.

 

[TJ Anderson]

There's no need to link L-N to IR test to earth for fault finding. The linked L-N method of IR testing is for initial verification of a new/altered circuit, or periodic testing. Fault finding does not follow standard test procedures as required for the above, if you link L-N you will not know whether the fault is L-E or N-E

He is testing with loads connected so L&N are connected across the loads.

 

[westward10]

Who is He?

 

[TJ Anderson]

Who is He?

The OP. Have just noticed he has asked two separate questions ... Both have the same answer though lol

 

[westward10]

The OP. Have just noticed he has asked two separate questions ... Both have the same answer though lol

Sorry I have confused you with the OP.

 

[phil lancaster]

if n&e are joined and then tested to earth surely this this would not be poss surely I would need to test the n and e unjoined to get a resistance?

 

[TJ Anderson]

if n&e are joined and then tested to earth surely this this would not be poss surely I would need to test the n and e unjoined to get a resistance?

Yes. He meant test between L&E and then N&E as seperate tests as opposed to L linked to N & tested to E.

 

[TJ Anderson]

.

 

[phil lancaster]

Ok thank you all for your input will let you know what happens

 

[Lucien Nunes]

What a lot of confusingness.

1. Testing IR to find a leakage fault does not require L-N test, which in any case you should not do (and cannot get a meaningful result from) when loads are connected.

2. When testing to earth, the loads will usually link L & N together by a low resistance so the readings of L-E and N-E tests will be the same. Therefore, there is no point doing these two tests separately with loads connected.

3. Just in case the only parts of loads connected between L & N are not of low resistance and cannot stand 500V without damage, you should solidly connect L & N to prevent that any voltage appearing between them. Therefore, only one test can usefully and safely be done, which is from L & N linked, to E. This is how a PAT tester tests insulation.

4. Most devices can withstand 500V on this test, or put another way, if they can't they aren't truly safe. But any appliances or cables with surge protection will give a failure reading, usually something under one megohm, as the surge protection clamps the 500V test voltage. They will usually pass easily at 250V, so that is a good reason to use this voltage when doing a global test to localise a bad load device.

5. Tracing a fault on the installation itself, with all loads disconnected, can and should be done L-E and N-E separately at 500V.

Not all leakage is due to poor IR but it is a good place to start.

 

[Bob Geldoff1234]

If your IR tests are showing good then don't rule out a faulty RCD and/or a loose connection somewhere arcing.This seems to trip RCD's as well but don't know why.Also don't rule out damp and slugs which can come and go which is great fun when your testing the following day with no faults showing

 

[Gavin John Hyde]

Hi I know that insulation resistance testing is normaly carried out with the LN joined but I am always all ways nervous that some damage may sill be done as there is so much sensitive equipment around LEDs Computer eqipiment
phil

If you do the IR test at 250v its not much more than what you get from a good supply anyway, I regularly get 245 to 247v in Bath on some properties. so we are talking a couple of extra volts.

 

[Wilko]

Hi - Don't forget it's a DC voltage during the IR test, vs +/- 400V (peaks) when its energised.

 

[telectrix]

isn't the applied voltage dependent on how fast you crank the handle on a megger IR tester?

 

[Lucien Nunes]

No, provided you crank fast enough there's a governor with a slipping clutch to keep the generator rpm and voltage constant.

 

[PEG]

What a lot of confusingness.

1. Testing IR to find a leakage fault does not require L-N test, which in any case you should not do (and cannot get a meaningful result from) when loads are connected.

2. When testing to earth, the loads will usually link L & N together by a low resistance so the readings of L-E and N-E tests will be the same. Therefore, there is no point doing these two tests separately with loads connected.

3. Just in case the only parts of loads connected between L & N are not of low resistance and cannot stand 500V without damage, you should solidly connect L & N to prevent that any voltage appearing between them. Therefore, only one test can usefully and safely be done, which is from L & N linked, to E. This is how a PAT tester tests insulation.

4. Most devices can withstand 500V on this test, or put another way, if they can't they aren't truly safe. But any appliances or cables with surge protection will give a failure reading, usually something under one megohm, as the surge protection clamps the 500V test voltage. They will usually pass easily at 250V, so that is a good reason to use this voltage when doing a global test to localise a bad load device.

5. Tracing a fault on the installation itself, with all loads disconnected, can and should be done L-E and N-E separately at 500V.

Not all leakage is due to poor IR but it is a good place to start.

Cannot believe you did not get a like,for a concise and useful explanation...

You should have been in the Poseidon Adventure ....could have been an hour shorter