help with rcd problem

[Mr Mark Sparks]

evening all,
went to look at a mates cu the other day tt system 30ma rcd main switch tripping out randomly
earth loop to stake 150 ohms.
earth loop with everything connected 1.30 ohms.
before i did any insulation tests i did a full set of rcd tests and it wouldnt trip via my tester (mft1552) however test button did work,

so said i would change cu as replacement main switch for old cu not available,
replaced cu today with wylex dual rcd cu tested circuits as i put them back.

now the odd thing there is one lighting circuit that tests >299 l-n >299 l-e >100 n-e when i connect this it stops the rcd from tripping on test.
swapped rcds round and tried circuit on both side of cu but with this circuit connected rcd will not trip, with it removed it will.

what sort of issue can stop the rcd from tripping

for those who dont know me i am 22yrs time served with 2391



at the mo the circuit is just via mcb not on either rcd
and i am waiting to here if random tripping has stopped

thanks mark

 

[sparks1973]

at what point on the install did you test the RCDs...at the RCD itself...or further downstream...at a socket perhaps....its just i1`v heard of RCBos that wont test upclose to em...but a bit further downstream...

 

[Mr Mark Sparks]

i tested at various points throughout the house and the result was always the same with the lighting circuit connected they wont trip.
as you say i know that memshield 2 rcbos wont trip properly up close

 

[Des 56]

I wonder if there are lots of transformers in the circuit,I'm thinking that somehow it is replicating no trip test meter operations, by somehow saturating the circuit with Dc

 

[Jimmy Boy]

With the L/E/N on the light circuit removed from board did you do a continuity and ohms measurement test from the disconnected ends of the circuit to L/E/N on the board just to double check there isn't a interconnection ?
J

 

[oldtimer]

Why dont you connect the light circuit switch all the lights off at the switch then test the rcd if this works then move on to switching each individual light on until you get the same problem.

 

[Mr Mark Sparks]

I wonder if there are lots of transformers in the circuit,I'm thinking that somehow it is replicating no trip test meter operations, by somehow saturating the circuit with Dc

yes there is some but not a huge number i did try to trip the rcd both loaded and unloaded

could you explain your comment as i dont understand

 

[Mr Mark Sparks]

With the L/E/N on the light circuit removed from board did you do a continuity and ohms measurement test from the disconnected ends of the circuit to L/E/N on the board just to double check there isn't a interconnection ?
J

in a word no i did try n bar to n bar to check for borrowed neutrals

 

[Mr Mark Sparks]

Why dont you connect the light circuit switch all the lights off at the switch then test the rcd if this works then move on to switching each individual light on until you get the same problem.

i did try with all lights on and all lights off still the same but will try again one by one as head was spinning by this point

 

[oldtimer]

i did try with all lights on and all lights off still the same but will try again one by one as head was
spinning by this point

Me thinks that if you switch each light off on the suspect circuit and the rcd should work then switch one on at a time and if you do track it down to a section that has transformers on it then this needs to be highlighted to the manufacturer and even safety standards as there could be a lot of rcds up the country that will not trip and that aint good.

Plus have you thought about writing to the rcd manufacturer I take it , it was the same make and model you replaced ?

 

[Mr Mark Sparks]

Plus have you thought about writing to the rcd manufacturer I take it , it was the same make and model you replaced ?

contactum taken out and new wylex dual board fitted house built mid 80s and suspect its always been like it

 

[Guest55]

Lets deal with this in a logical manner.
So , did you carry out the rcd tests at the CU with all mcbs switched off ?
Did the both rcds test ok with regards to trip times ?
If they did then we know the problem is not the rcds.

 

[sparks1973]

I wonder if there are lots of transformers in the circuit,I'm thinking that somehow it is replicating no trip test meter operations, by somehow saturating the circuit with Dc

But that was D-loc....the Robin testers used it...no good on modern RCDs Des..as they will detect D-loc.....

 

[Guest55]

yes there is some but not a huge number i did try to trip the rcd both loaded and unloaded

could you explain your comment as i dont understand

Most gadgets and appliances create "electronic noise" on a circuit which can affect rcd testing - which is why mcbs should be switched off at the time of testing.

 

[sparks1973]

look....i think its now clear that something on that lighting circuit is causing the RCDs not to let go....so now its been narrowed down to a final circuit...lets start to look at that.....i mean this is standard fault finding practice surely guys...

 

[sparks1973]

i`m wondering if its a problem being caused by a faulty Dimmer switch....or if i`m just being a dimmer?

 

[Plonker 3]

Most gadgets and appliances create "electronic noise" on a circuit which can affect rcd testing - which is why mcbs should be switched off at the time of testing.

My ELECSA assessor, told me to test at a socket or light even on a DB change where you might not be 100% certain of having everyting unplugged or disconnected.

 

[sparks1973]

it has to be something that is common to BOTH RCDs from this one lighting circuit for them both not to want to trip....that means the earth bar........

 

[Guest55]

.....i mean this is standard fault finding practice surely guys...

Nothing is straightforward or standard in trying to diagnose a fault over the internet.

 

[sparks1973]

My ELECSA assessor, told me to test at a socket or light even on a DB change where you might not be 100% certain of having everyting unplugged or disconnected.

yes...its better to be downstream of the device to test...rather than right on it...

 

[Guest55]

My ELECSA assessor, told me to test at a socket or light even on a DB change where you might not be 100% certain of having everyting unplugged or disconnected.

I couldnt give a rats --- what your assessor says lol.
And if all the mcbs are turned off it doesnt matter whats plugged in or not ???

:41:

 

[sparks1973]

Nothing is straightforward or standard in trying to diagnose a fault over the internet.

oh come on Biff....you should know me better than that by now...lol...

 

[Guest55]

yes...its better to be downstream of the device to test...rather than right on it...

Why is that ??

 

[Plonker 3]

yes...its better to be downstream of the device to test...rather than right on it...

But what about unknown loads that may contain a capacitor which would give false readings?

- - - Updated - - -

Why is that ??

On a new build then yes I would say so, but a refurb not possible really which was what I was referring too.

 

[Des 56]

I made the suggestion about equipment perhaps being the problem because,I trust Mark has carried out all the usual tests we would normally do,he has happened upon an unusual problem that requires an unusual cause

I made the reference to the no trip test of a loop tester to give comparison to what I was dreaming about

Dreaming,because I have no technical know how whatsoever, regards the intricacies or the workings of the rcd tester

I did however find this bit of information from a study into the effects

DC in Residual Current Devices
The widespread use of semi-conductor
devices in domestic, commercial and industrial equipment may result in the
flow of non-sinusoidal earth fault currents in the event of equipment insulation
failure. This can occur in certain circumstances where the waveform may be
rectified or chopped and these waveforms are said to contain a pulsating d.c.
component which may desensitise a standard Type AC RCD.

The above is somewhere along the lines of what I was thinking,it may be utter bullcarp and the problem will likely be standard, but its certainly an unusual cause


Some of the equipoment listed includes Phone chargers invertors,transformers etc

 

[sparks1973]

Why is that ??

because some devices dont test properly up close....

 

[sparks1973]

I made the suggestion about equipment perhaps being the problem because,I trust Mark has carried out all the usual tests we would normally do,he has happened upon an unusual problem that requires an unusual cause

I made the reference to the no trip test of a loop tester to give comparison to what I was dreaming about

Dreaming,because I have no technical know how whatsoever, regards the intricacies or the workings of the rcd tester

I did however find this bit of information from a study into the effects

DC in Residual Current Devices
The widespread use of semi-conductor
devices in domestic, commercial and industrial equipment may result in the
flow of non-sinusoidal earth fault currents in the event of equipment insulation
failure. This can occur in certain circumstances where the waveform may be
rectified or chopped and these waveforms are said to contain a pulsating d.c.
component which may desensitise a standard Type AC RCD.

The above is somewhere along the lines of what I was thinking,it may be utter bullcarp and the problem will likely be standard, but its certainly an unusual cause


Some of the equipoment listed includes Phone chargers invertors,transformers etc

hey no probs Des...i wasn`t having a go mate....it was just a bit of info i picked up some time ago about the old robin D-loc...thats all mate...

 

[Guest55]

because some devices dont test properly up close....

Only mem rcbo's have that problem , most other manufacterers advise the testing to be done at the output terminals.

 

[Jimmy Boy]

The OSG does advise testing as near as practicable to it's point of installation, and load supplied should be disconnected, but struggling to find that in the BGB
J

 

[Mr Mark Sparks]

just to clarify as of now the board has 4 circuits on one rcd and these all test and behave normaly

3 circuits on the other rcd and they also test and behave normally

the problem circuit is just on mcb at mo but would effect either rcd if connected

have just had text to say nothing has tripped since i left at 1pm and on the old board it was going upto 10 times a day

so i am wondering if i have a transformer or something that tested ok cold after cu change but is stopping rcd tripping and also when it gets hot causes insulation fault remember l-e 299 n-e 100

 

[Plonker 3]

Could you not put the circuit you "think2 is causing the problem on a RCBO, especially as it is a TT system?

 

[Guest55]

The OSG does advise testing as near as practicable to it's point of installation, and load supplied should be disconnected, but struggling to find that in the BGB
J

It should be in GN3 , gives more detail on the nuts and bolts of testing.

 

[Mr Mark Sparks]

Could you not put the circuit you "think2 is causing the problem on a RCBO, especially as it is a TT system?

yep going to put one in tomorrow to see if it will trip under test or nuisance trip.

 

[Plonker 3]

Either way, I would leave it in once fitted.

 

[sparks1973]

11.1 osg p 106...so yes..i was wrong Biff on the testing too close thing...although some devices (MEMs) dont like it...

 

[Nath88]

I remember going to a job a collegue had been to..... Rcd main switch not tripping so he ordered a new one and i went to fit it on his day off.
When i went to test it, wouldnt trip either so i did my tests and found a low resistance fault between Neutral and earth which turned out to be a trapped neutral in the fused spur. So sorted that and then checked both rcds and both were fine.
Just an idea

 

[sparks1973]

Could you not put the circuit you "think2 is causing the problem on a RCBO, especially as it is a TT system?

that may just mask the problem...not solve it as who says the RCBO will function?..

 

[Jimmy Boy]

11.1 osg p 106...so yes..i was wrong Biff on the testing too close thing...although some devices (MEMs) dont like it...

Thats the one !
J

 

[Plonker 3]

that may just mask the problem...not solve it as who says the RCBO will function?..

Maybe but at least the circuit would then comply with the regs in regards too TT systems.

 

[sparks1973]

Maybe but at least the circuit would then comply with the regs in regards too TT systems.

but if the RCBO doesnt want to let go....well then it wont comply will it..