UK 100mA RCCB keeps tripping with certain items connected.

[Jackster]

Hi all

I am in an industrial unit and the 100mA RCCB keeps tripping.
They use it as the main switch inside of the unit's distribution board. The mainboard that feeds all the units has an MCB in it.

The RCCB keeps tripping when I plug-in AC to DC transformer loads such as laptops, speakers, monitors etc.
The 30mA RCBO that does the sockets (and my lights in my office) does not trip.
And I don't have all of these devices plugged in at the same time. Right now my laptop is plugged in with the other office equipment, but my monitors and speakers are not. If I plug those in, it trips. If I put any load on my constant current/voltage DC power supply, it trips. Talking less than 1w and it trips.

I have already replaced it with another RCCB of the same make and model just to see if the old one is faulty but it does the same thing.
When I swapped the RCCB over, it tripped as soon as I connected it back up to the board. The MCB was off, though it was only a 3 pole...

It also trips randomly when everything has been fine for hours. Not much if anything connected and it will just trip.
Every night it trips.

Any ideas on what I can do? What tests can I do?
I am thinking it is a natural leaking to earth somewhere? I should check between ground and neutral on the incoming supply and within my unit?

Thanks
Jack

(yes I am aware that I should get an electrical in, the site electrician was meant to be on site last week but never showed and I will be getting my personal electrical in next week or so when they can book me in)

 

[HappyHippyDad]

Hi Jackstar.
You just need to IR the consumer unit with a multi function tester, which may or may not show a fault. If it shows a fault you just track it down by breaking down the faulty circuit and IR testing.

If it doesn't show a fault then this points to natural leakage which you confirm with an earth leakage tester around the main tails. Your electrician should have those 2 tools. If its the latter then change the board for an RCBO board, so each separate breaker takes a smaller amount of earth leakage.

Ps.. I am purely domestic, so may have missed some other basic steps involved with industrial.

Pps.. of course it could just be one item of equipment causing the problem!

 

[timhoward]

I am thinking it is a natural leaking to earth somewhere? I should check between ground and neutral on the incoming supply and within my unit?

I agree it sounds like a N to E fault.
Don’t test the incoming supply - N is always joined to E somewhere, which might be inches away in your supply cut out or might be back at the substation transformer.

Does the test button work on your 30ma RCD , proving that works would be high on my list to be sure which side of it the real fault is.

In the case of N/E faults the connected load can be driving current through a fault on a completely different circuit so it’s almost impossible to make progress without an IR tester, an RCD tester and a ma clamp meter.

It sounds like the issue is on another circuit, upstream of the 30ma RCD but downstream of the 100ma RCD.
If you are struggling to get someone around you could mention what area you are in.

 

[pc1966]

I agree it sounds like a N to E fault.

That sounds likely, though the sensitivity to small things plugged in is odd.

Don’t test the incoming supply - N is always joined to E somewhere, which might be inches away in your supply cut out or might be back at the substation transformer.

If the incomer is a RCCB then it ought to cut the neutral as well, even if 3-phase, so IR testing L+N to E would show insulation faults.

Does the test button work on your 30ma RCD , proving that works would be high on my list to be sure which side of it the real fault is.

This!
Very important for the OP to check is if you have a fault on the sockets that the 30mA is simply not responding to then it would be the incomer that goes next.

In the case of N/E faults the connected load can be driving current through a fault on a completely different circuit so it’s almost impossible to make progress without an IR tester, an RCD tester and a ma clamp meter.

It sounds like the issue is on another circuit, upstream of the 30ma RCD but downstream of the 100ma RCD.
If you are struggling to get someone around you could mention what area you are in.

Hopefully they get someone soon to look at it!

 

[timhoward]

If the incomer is a RCCB then it ought to cut the neutral as well, even if 3-phase, so IR testing L+N to E would show insulation faults.

Yeah thing is I was a little concerned that he might start using non gs38 probes on the bottom of the incoming device - it was a little ambiguous what he meant by testing the incoming supply! Safety first and all that….

 

[pc1966]

Yeah thing is I was a little concerned that he might start using non g38 probes on the bottom of the incoming device - it was a little ambiguous what he meant by testing the incoming supply! Safety first and all that….

Yes, very good point! Getting a non-CAT-IV meter on the wrong range there, or accidentally touching a probe, is probably going to end very badly.

 

[Jackster]

Yeah thing is I was a little concerned that he might start using non gs38 probes on the bottom of the incoming device - it was a little ambiguous what he meant by testing the incoming supply! Safety first and all that….

Yes, very good point! Getting a non-CAT-IV meter on the wrong range there, or accidentally touching a probe, is probably going to end very badly.

My £20 CAT4 meter from eBay is very good. Doubt it will explode. It is Fluke yellow


Thanks for all the tips guys. I have a multi function meter arriving tomorrow on loan from Amazon and will do everything suggested here.

I just remembered that I removed the neutral fly lead from the socket RCBO the other day and it stopped tripping the RCCB.
Pretty sure I tested the RCBO button but I will test again later and report back.

 

[timhoward]

Thanks for all the tips guys. I have a multi function meter arriving tomorrow on loan from Amazon and will do everything suggested here.

For the avoidance of doubt, speaking for myself, I was not suggesting you start diving in with a tester!

 

[westward10]

It is wholly inappropriate for you to obtain a tester and then start testing the circuits. As competent people we understand what, how and why we are testing and in a safe manner, we can also interpret the results we are getting. I would recommend you wait until an electrician can attend your site.

 

[Jackster]

Just pressed the test button on the sockets RCBO and it tripped. The main RCCB did not trip.

 

[pc1966]

Just pressed the test button on the sockets RCBO and it tripped. The main RCCB did not trip.

That suggests the fault is not actually on the circuit supplied by that RCBO. However, as many RCBO don't isolate the neutral (unlike RCCB that are normally all-pole switching) then it could still be a N-E fault downstream of the RCBO.

 

[HappyHippyDad]

That suggests the fault is not actually on the circuit supplied by that RCBO. However, as many RCBO don't isolate the neutral (unlike RCCB that are normally all-pole switching) then it could still be a N-E fault downstream of the RCBO.

That's narrowed it down then

 

[pc1966]

That's narrowed it down then

Yes, it is not really that helpful unless the details of the RCD device are known, if it is N-switching or not.

But at least it identifies the fault is not on the L-side!

 

[Jackster]

Yes, it is not really that helpful unless the details of the RCD device are known, if it is N-switching or not.

But at least it identifies the fault is not on the L-side!

When I was swapping out the RCCB, I also disconnected that RCBO's neutral and it did not trigger the RCCB. But as soon as I touched the lead to the neutral bar, it tripped.
What is odd about it all though. I can have nothing plugged in and come back the next day to it being tripped.

 

[pc1966]

When I was swapping out the RCCB, I also disconnected that RCBO's neutral and it did not trigger the RCCB. But as soon as I touched the lead to the neutral bar, it tripped.
What is odd about it all though. I can have nothing plugged in and come back the next day to it being tripped.

N-E faults are like that. Often there is very little difference in N & E potential (that is almost the definition of neutral, as it is linked to earth somewhere) so no trip.

But then something draws a lot of current (and it could be external to the installation) thus causing N to change due to the volt drop and that allows enough stray to E for the RCCB to trip!

 

[pc1966]

Actually thinking a bit more - if there is a N-E fault downstream of the 30mA RCBO it should also trip as it would see the imbalance as well. That assumes the trip speeds are selective, is the 100mA one a delay RCCB (i.e. "type S" for selective)?

 

[Jackster]

Actually thinking a bit more - if there is a N-E fault downstream of the 30mA RCBO it should also trip as it would see the imbalance as well. That assumes the trip speeds are selective, is the 100mA one a delay RCCB (i.e. "type S" for selective)?

I guess we shall find out when my "incompetent" --- gets a multi-function tester on Monday

 

[timhoward]

I guess we shall find out when my "incompetent" --- gets a multi-function tester on Monday

Don’t take it personally! We have to assume that unknown people can’t test safely. In colleges testing is taught after installation and after experience has been built up with how things are typically wired.
And electrical testing can be a dangerous business and doesn’t always give clear answers - there are regular questions on here from qualified sparks about interpreting results. Some tests are dead and some are live and it’s very easy to make a mistake.
We just care about safety.

 

[Jackster]

Well, I am still alive.

So I managed to find my clamp meter and it is showing 0.15A to 0.5A on the 20A setting depending on if I clamp near the RCBO or the main tails where the 4mm cable is accessible.
Don't think one can trust it... Ideally need something that measures in the 2A range. Will be worth it in my electronics lab so I will grab one tomorrow.


Got my hands on the multifunction tester. A pretty good bit of kit to have a play with. Found a fault on an old extension cord (not been used in years) so that has gone in the bin so already paying itself off.

I did the RCD tests on both AC/AC+DC settings. The 30mA RCBO for the sockets was fine as was tripping within spec.
The 100mA RCCB tripped at a max of 105mA on AC and at 450mA on AC+DC. Not 100% sure about those results.


I did an IR test on the board between N+E and L+E * 3. All fine.

Having been messing around with all the plugs and sockets (mostly extension cables in the office area), I have reduced the apparently N to E leakage down from the average reading on the socket circuit of 0.4-0.5A to 0.1-0.2A.
Put a few new 8 way extensions in while I was down under the desks which might have helped?

I did do an IR test on all the existing multi-way extension cables and all were fine on 500v.

The RCCB has yet to trip under normal conditions but as soon as I put a load on my benchtop AC to DC power supply, it trips.
Could be a fault with that bit of kit but that kit is not normally plugged in. I tested it at home and the RCD did not trip there.

 

[pc1966]

Well, I am still alive.

A good start to the day!

So I managed to find my clamp meter and it is showing 0.15A to 0.5A on the 20A setting depending on if I clamp near the RCBO or the main tails where the 4mm cable is accessible.
Don't think one can trust it... Ideally need something that measures in the 2A range. Will be worth it in my electronics lab so I will grab one tomorrow.

Something like this is needed:

https://cpc.farnell.com/multicomp-pro/mp780050/mini-leakage-current-tester/dp/IN08364

Also you need to get the L and matching N in to the clamp to see the residual. I would strongly advise against moving cables in an energised DB! Try and get anything to measure in-place and accessible with all power off.

Got my hands on the multifunction tester. A pretty good bit of kit to have a play with. Found a fault on an old extension cord (not been used in years) so that has gone in the bin so already paying itself off.

I did the RCD tests on both AC/AC+DC settings. The 30mA RCBO for the sockets was fine as was tripping within spec.

Good to hear.

The 100mA RCCB tripped at a max of 105mA on AC and at 450mA on AC+DC. Not 100% sure about those results.

If it is type AC and tested on a type A waveform with DC component then it can be desensitised.

If 100mA rating it ought to trip between 50-100mA. You need to have all loads off so you are just testing it alone, and not having system currents also present.

I did an IR test on the board between N+E and L+E * 3. All fine.

Having been messing around with all the plugs and sockets (mostly extension cables in the office area), I have reduced the apparently N to E leakage down from the average reading on the socket circuit of 0.4-0.5A to 0.1-0.2A.
Put a few new 8 way extensions in while I was down under the desks which might have helped?

Those are massive currents! Is that measured with L&N in the clamp, or are you (incorrectly) clamping the CPC? If so then any metalwork anything is in contact with could be causing odd currents to circulate.

I did do an IR test on all the existing multi-way extension cables and all were fine on 500v.

OK.

The RCCB has yet to trip under normal conditions but as soon as I put a load on my benchtop AC to DC power supply, it trips.
Could be a fault with that bit of kit but that kit is not normally plugged in. I tested it at home and the RCD did not trip there.

Have you IR tested the power supply? I.e. the sort of basic PAT testing part (along with CPC bond resistance, etc)

Are you sure the power supply is going in to the RCBO's supply? There is something inconsistent with being able to trip the 100mA incomer and not trip the 30mA socket feed, unless the installation happens to be sitting only a few mA leackage below the trip point normally.

 

[Jackster]

Something like this is needed:
https://cpc.farnell.com/multicomp-pro/mp780050/mini-leakage-current-tester/dp/IN08364
Also you need to get the L and matching N in to the clamp to see the residual. I would strongly advise against moving cables in an energised DB! Try and get anything to measure in-place and accessible with all power off.

Ah nice find. The one I was looking at was £5 more and not as accurate.

I am doing the L + N matching pair for the sockets and the DB. Both are accessible without moving the cables.

If it is type AC and tested on a type A waveform with DC component then it can be desensitised.

If 100mA rating it ought to trip between 50-100mA. You need to have all loads off so you are just testing it alone, and not having system currents also present.

I will test again on Wednesday when I am back at the unit. Or tomorrow night.

Those are massive currents! Is that measured with L&N in the clamp, or are you (incorrectly) clamping the CPC? If so then any metalwork anything is in contact with could be causing odd currents to circulate.

I am clamping the Live and Nutrual 4mm singles coming out of the RCBO. It ranged from 0.2A to 0.7A but it really came down to how I was holding the clamp meter on the cables and at what angle. That is why I don't believe my 20A clamp meter. Will wait for the one from CPC to arrive.

Have you IR tested the power supply? I.e. the sort of basic PAT testing part (along with CPC bond resistance, etc)

Are you sure the power supply is going in to the RCBO's supply? There is something inconsistent with being able to trip the 100mA incomer and not trip the 30mA socket feed, unless the installation happens to be sitting only a few mA leackage below the trip point normally.

I tested N+E and L+E on that power supply up to 500v. No issues. Will look into the PAT procedure and get back to you on that.


There are only 3 circuits on the DB. Sockets, Door and Lights. I can follow the 4mm singles from the sockets up into the RCBO.
This is the reason why I am struggling to get my head around it all.

 

[pc1966]

There are only 3 circuits on the DB. Sockets, Door and Lights. I can follow the 4mm singles from the sockets up into the RCBO.
This is the reason why I am struggling to get my head around it all.

Check the general leakage to see if it is close to the 100mA RCCB's trip threshold.

Check that the 100mA RCCB is a delay type (S), otherwise a very short spike might just trip it before the 30mA one goes.

 

[Jackster]

Check the general leakage to see if it is close to the 100mA RCCB's trip threshold.

Check that the 100mA RCCB is a delay type (S), otherwise a very short spike might just trip it before the 30mA one goes.

It has an S on the front off it.

When I checked using my crappy clamp meter, it said nearly 1 amp...

I did test the meter and it was saying I was drawing 11A during normal use of the office.
That is a 1700w blower heater, 800w (max) IR panel, 2 monitors, laptop, printer, 160w of LED tubes and a little bit of networking gear.. So somewhat accurate there.

Just noticed the power has gone out at the unit. Either the internet has gone (doubt it) it the power tripped.

 

[Jackster]

The half-decent leakage current tester arrived today and it is very good. Very nice build quality for the price.

So I tested a load of stuff and my space heater, is leaking around 60ma. My LED lights/printer/network/IR heater circuit is leaking around 30ma.

Without these two circuits, my leakage is down in the 3-5mA range.
With these circuits, it is in the 100-125mA range for the whole DB.

So I guess we have found the issue. As to why the RCCB trips and not the RCBO, no idea.
The RCBO is 30mA and the RCCB is 100mA.
I tested the RCBO and it trips fine on the Megga multifunction tester. The RCCB trips at 110mA.

I have done an IR test on the space heater and it is fine up to 500V.
Think I will retire that anyway as it is on recall apparently...

Will go through the other circuit at a later date to see what in that is causing the 30mA leakage.

 

[DPG]

The half-decent leakage current tester arrived today and it is very good. Very nice build quality for the price.

So I tested a load of stuff and my space heater, is leaking around 60ma. My LED lights/printer/network/IR heater circuit is leaking around 30ma.

Without these two circuits, my leakage is down in the 3-5mA range.
With these circuits, it is in the 100-125mA range for the whole DB.

So I guess we have found the issue. As to why the RCCB trips and not the RCBO, no idea.
The RCBO is 30mA and the RCCB is 100mA.
I tested the RCBO and it trips fine on the Megga multifunction tester. The RCCB trips at 110mA.

I have done an IR test on the space heater and it is fine up to 500V.
Think I will retire that anyway as it is on recall apparently...

Will go through the other circuit at a later date to see what in that is causing the 30mA leakage.

That's a very high leakage. Is that meter accurate down to low mA figures?

 

[pc1966]

It is odd to see 100mA leak and no DC effect, unless there is some largish capacitor in there! Or maybe something odd happens when it gets hot. Either way, that is a rather sick heater!

As you are running close to the up-front RCCB's threshold the trips make perfect sense. Without the heater can you now plug in stuff, etc, and all is happy?

 

[Jackster]

It is odd to see 100mA leak and no DC effect, unless there is some largish capacitor in there! Or maybe something odd happens when it gets hot. Either way, that is a rather sick heater!

As you are running close to the up-front RCCB's threshold the trips make perfect sense. Without the heater can you now plug in stuff, etc, and all is happy?

I will remove the heater from the circuit when I leave and see if it trips overnight. I have removed all the plugs from the sockets before and have had it trip so I don't know if this is the only issue here.

 

[pc1966]

I will remove the heater from the circuit when I leave and see if it trips overnight. I have removed all the plugs from the sockets before and have had it trip so I don't know if this is the only issue here.

OK.

Did you test the heater's whole circuit (i.e. IR it at the CU) or just disconnected the heater to check it alone? Might be nasty cable fault developing otherwise.

 

[Jackster]

It just tripped again. I removed the heater and it still trips. My lights/printer/IR heater (turned off) and network circuit is fine. <5mA leak.

But as soon as I plug in my office equipment (Laptop, speakers, monitors, phone etc) and it trips. These were only adding 5mA earlier.

https://drive.google.com/file/d/1KQiSVuwgiJi9dorh8_jdm3vHrAHscI0G/view?usp=sharing

(should be able to view shortly without downloading)

Literally just as I am about to post this, the site spark at door.
He will be around tomorrow to see what's up.

 

[pc1966]

But as soon as I plug in my office equipment (Laptop, speakers, monitors, phone etc) and it trips. These were only adding 5mA earlier.

Have you opened that socket to inspect it? Or try a new one?

It might be the actual act of inserting the plug is causing a fault, maybe N-E so no bang but enough to trip the RCCB.

 

[pc1966]

If you plug in a big load there, say a heater, then switch the heater on without physically disturbing the socket, is it OK?

Trying to work out if some odd neutral path or similar. I'm guessing you have IR tested that circuit to rule out N-E permanent fault?

 

[Jackster]

OK.

Did you test the heater's whole circuit (i.e. IR it at the CU) or just disconnected the heater to check it alone? Might be nasty cable fault developing otherwise.

I did an IR test on it up to 500v and there was no fault. I did it while it was warm also.

Have you opened that socket to inspect it? Or try a new one?

It might be the actual act of inserting the plug is causing a fault, maybe N-E so no bang but enough to trip the RCCB.

I will test but it tripped on other sockets too. That socket is only 6 months old.

If you plug in a big load there, say a heater, then switch the heater on without physically disturbing the socket, is it OK?

Trying to work out if some odd neutral path or similar. I'm guessing you have IR tested that circuit to rule out N-E permanent fault?

Will try that. I have IR tested the socket circuit and no issues.


Hopefully, the electrician can work it out tomorrow.

 

[pc1966]

I did an IR test on it up to 500v and there was no fault. I did it while it was warm also.

OK.

I will test but it tripped on other sockets too. That socket is only 6 months old.

Have you PAT tested what you are plugging in? Maybe faulty plug/adaptor and strain/movement is causing it to N-E short?

I am still puzzled by why the 30mA RCBO is not tripping in that case though.

Will try that. I have IR tested the socket circuit and no issues.

Hopefully, the electrician can work it out tomorrow.

Hopefully!

 

[nicebutdim]

That's a very high leakage. Is that meter accurate down to low mA figures?

If it's the one from CPC, linked by @pc1966, I can confirm that they're fairly accurate - on a par with other more expensive testers.

 

[Jackster]

So I am hearing this second hand as I was not here, but the spark looked at the DB and the unit's plumed in circuits and they are all okay.
He has stated that I "need to get my equipment PAC tested" and "my equipment is tripping the breaker".

 

[DPG]

So I am hearing this second hand as I was not here, but the spark looked at the DB and the unit's plumed in circuits and they are all okay.
He has stated that I "need to get my equipment PAC tested" and "my equipment is tripping the breaker".

He would have said 'PAT tested'. Which is what you were going to do back in post #21. Seems like a good thing to get done.

 

[pc1966]

Am I doing this right? This result expected?

Looks right method, results are bizarre!

The 1mA on the CPC would be reasonable for a floating bit of kit and some filter capacitance.

The 39mA for the L-N difference is difficult to explain and ought to be triggering the RCBO that feeds it.

What is attached to the cable you are testing, has it any external cables such as network or RS232, etc, or is in contact with building metalwork, machines being controlled, etc?

I am still going through all the kit in the office doing IR, earth to case and polarity check. No issues so far.

If you IR test that kit you clamped (above) what do you get?

And when doing the IR test try having the CPC/earth also linked to the supply earth at the time.

Edit to add: Have you used your MTF to ramp-test the RCBO from that socket that you are getting 39mA and no trip from?

 

[Jackster]

Looks right method, results are bizarre!

The 1mA on the CPC would be reasonable for a floating bit of kit and some filter capacitance.

The 39mA for the L-N difference is difficult to explain and ought to be triggering the RCBO that feeds it.

What is attached to the cable you are testing, has it any external cables such as network or RS232, etc, or is in contact with building metalwork, machines being controlled, etc?

If you IR test that kit you clamped (above) what do you get?

And when doing the IR test try having the CPC/earth also linked to the supply earth at the time.

Edit to add: Have you used your MTF to ramp-test the RCBO from that socket that you are getting 39mA and no trip from?

Thought I was not losing it...

That kit is a small 2kW space heater. The RCBO never trips unless I use the Megga's RCD test function or the button on the RCBO.
It is a full plastic housing space heater and the RCCB trips with or without it.

IR tested it and it is fine when powered off. Obviously if I IR L+N, when powered on it, is a resistive short.

Ramp testing using that socket, 21mA


Going to leave things for now and see if it trips again. It was fine yesterday after tripping a few times when I plugged in my computer equipment. It is all still connected but is not tripping. I do have a 5kW 3ph load cycling on and off though so not sure if that is helping prevent it from tripping?

 

[pc1966]

Have you tired taping the L & N on your test plug/socket together and checking the L-N clamped current again?

Have you tried clamping the whole 3-core cable?

Usually the fact both conductors are through the clamp is enough, but you can see some imbalance if they are quite different paths.

 

[Jackster]

PAT guy came over and tested everything. All passed. Even my work passed.

 

[Jackster]

Site electricians came today and swapped the RCCB for an isolator.

Problem solved!

 

[westward10]

@Jackster what you have done may well have compromised the safety of the installation but if an electrician carried out this 'remedy' then I would assume they feel otherwise and this may be the case. On that note this thread is closed.