RCD Problems. Serious Head Scratching!!

[BangTidy]

Loooking for a bit of ideas on a issue if have with a RCD. 4 circuits on a submain that i recently changed, above the board was a leak so intally thought the problem was related to that. When carrying out the RCD auto test when it came to x1 the meter read error and would not trip. If i set it to x5 it tripped.

So put it down to a faulty RCD and changed it today, so when i auto tested this time it only goes and trips on x 1/2.
First off i checked my meter on another circut testing an RCBO on the main board which went fine.
So i disconnected all the loads on the problem board and tested individually but still the problem of trip on x 1/2. Next i disconnected the incoming supply and carried out a insulation resistance test 500v between conductors, no problem there.

I carried out a insulation test on the lighting circuit and had a reading of 0.04M between Live and Neutral but with all the lights as GU10 fluorescents would i be right in thinking the reading is down to this?
Anyway with the lighting circuit disconnected from the board still the RCD trips on x 1/2 so this has got me seriously scratching my head.
What i wondering is the board is the 3rd submain from the main panel and there is a 2 way RCD unit supplying a outside socket in the middle which im unsure where is fed from at this stage. Could this be a problem having another RCD downstream?

In the end put the old RCD back in tested that and is ok on x 1/2 but get the error message and no trip on x 1 again!!

Came away to get my thinking cap on and any help on what the problem may be?
The tester is a megger 1553 and was calabrated 6 months ago.

Any ideas gratefully accepted?

Cheers

 

[telectrix]

are you sure you'd set the 1553 on the correct mA range? if so, sounds like faulty RCDs.

 

[topquark]

What he said ^^^^^^ sounds like it's maybe set on 10mA range.

 

[Kamikaze]

I have had this before and eventually tracked it down to a high resistant neutral joint on the circuit. Check the voltage at the end of the circuit and see if it’s what you are expecting. If I’m right you should get some sort of a voltage reading between neutral and earth.

 

[BangTidy]

I have had this before and eventually tracked it down to a high resistant neutral joint on the circuit. Check the voltage at the end of the circuit and see if it’s what you are expecting. If I’m right you should get some sort of a voltage reading between neutral and earth.

To the other replys, no had my meter set to 30ma so its not that.

So it could be a loose neutral somewhere on a circuit?
Theres 4 circuits, 2 rings, 1 radial and a lighting circuit. The lighting theres about 20 GU10 on that circuit.
With my meter i did have it set at times to voltage on the power circuits and it showed 234v. Should i be looking to test every point then?

 

[Des 56]

So i disconnected all the loads on the problem board and tested individually but still the problem of trip on x 1/2.

Are you testing the rcd independant at source ?

 

[BangTidy]

Its a commercial premises so theres hardly any RCD "at source" if you mean the main board. Theres only 1 RCBO on a dedicated circuit on the main panel.

 

[Engineer54]

Do you have a ''known'' to be good RCD that you can try?? Your getting 2 different faults on 2 different RCDs, which points to 2 duff RCDs, ...and that i'm afraid to say is not an uncommon occurance. One of the reasons i don't not put total trust in the things!! ...lol!!

 

[topquark]

So i disconnected all the loads on the problem board and tested individually but still the problem of trip on x 1/2.

Are you testing the rcd independant at source ?

Its a commercial premises so theres hardly any RCD "at source" if you mean the main board. Theres only 1 RCBO on a dedicated circuit on the main panel.

I think Des is suggesting testing at the load side of the RCD, so you can eliminate the final circuit wiring of the four circuits (and any associated poor connection) from your enquiries.

 

[Kiers1970]

Had a similar fault this week, found a switch to be tightened against a strapper behind a switch. Prob wont be this in this instance but learnt something this week, RCD's under test that will not operate usually indicates a fault on the final circuits its supplying. And always to test your 2 way circuits !

 

[Des 56]

I will agree rcds can and do fail
I will also say that in the majority of cases with these types of problems,the problem invariably gets nailed down to installation faults

There are occasions when testing rcds on a circuit fed by that rcd, that tripping of the rcd becomes eratic with times, or even no trip, when the rcd itself is not the problem,the problem is sometimes the method of testing or even the test meter itself

 

[Deleted member 26818]

Do you have a ''known'' to be good RCD that you can try?? Your getting 2 different faults on 2 different RCDs, which points to 2 duff RCDs, ...and that i'm afraid to say is not an uncommon occurance. One of the reasons i don't not put total trust in the things!! ...lol!!

My thoughts exactly.

 

[Kamikaze]

To the other replys, no had my meter set to 30ma so its not that.

So it could be a loose neutral somewhere on a circuit?
Theres 4 circuits, 2 rings, 1 radial and a lighting circuit. The lighting theres about 20 GU10 on that circuit.
With my meter i did have it set at times to voltage on the power circuits and it showed 234v. Should i be looking to test every point then?

It wasn’t a loose connection. Water got into the connector and corroded the joint. As soon as I read your original post I had flashbacks to that particular job but that does not mean your problem is the same. In my case it had me and two other electrician going for a good few hours…lol. All the insulation resistance and continuity tests were ok but what gave it away was this random voltage which would appear and disappear. Initially put this down to the testers playing up but when we tried other testers, they were also doing it. Still can’t get my head round why that would stop the RCD from tripping but I’m sure some of the guys will put me straight on that…lol

 

[tony mc]

IMO if useing a Megger 1553 I would say the new RCD is faulted on half test.

With all loads off the first thing I would have done is ramp test it to see how many mA it is tripping out.

The original RCD I would say is also faulted with the water damage IMO the connection for you live prob is unstable fluctuating voltage giving you a error code.

The Megger should give value of >400 ms on 1 x if it does not trip and >40 ms on 5 x test

Obviously the tests are as the others have said with no load on the RCD.

I would also have a look at the voltage on the load side of the RCD that might give you the answer!

 

[BangTidy]

Cheers for the replys, ive got hold of another RCD from the wholesalers and give that a try and see tomorrow. Ill put it through a ramp test aswell and get back with the outcome.
Hoping it was just another faulty RCD.

Cheers again for the ideas.

 

[Kiers1970]

Cheers for the replys, ive got hold of another RCD from the wholesalers and give that a try and see tomorrow. Ill put it through a ramp test aswell and get back with the outcome.
Hoping it was just another faulty RCD.

Cheers again for the ideas.

It wont be mate, guarantee you have a fault somewhere, like i said earlier, I had the same thing this week.

 

[Engineer54]

It wont be mate, guarantee you have a fault somewhere, like i said earlier, I had the same thing this week.

So which of the faults did you have this week then?? Because our guy here has experienced 2 different faults on two different RCDs. So if he has an installation fault, it's a fault that affects 2 RCDs at two different functional levels!!

I'm not saying that it isn't an installation fault, because RCD are, or can be fickle things.... But it is more likely that these two RCDs are faulty units...

 

[Kiers1970]

be a hell of a coincidence.

 

[Kiers1970]

this week this happened
a fellow electrician changed his consumer unit to a twin RCD CU
when using appliances in his kitchen, RCD tripped
this led him to believe logically that he had a faulty appliance/ wiring
tested wiring, ok, changed the kitchen ring over to the adj rcd, all ok.
so he as the OP wazs left head scratching
called me in.
reconnected and tested, rcd set no probs, all lights and power ok
used a few kitchen apps, rcd goes.
test 1st rcd, no probs
test second rcd, no trip.
all circuits test out fine!
so , disconnect all neutrals, switch off mcbs, rcd trips fine. within times.
go thru all circuits one by one, remove and replace neutrals as we go, rcd testing as I go , bingo, fault on lighting circuit, strapper crushed behind switch, made good, tested all ok.
OP , try this, I've not read if you've tried this, or if youve tried the rcd test with no circuits connected, if it operates fine without any circuits connected, its more than likely ok.

 

[Engineer54]

be a hell of a coincidence.

Not as uncommon as you may think Kiers, we've had, ...on more than occasion whole batches that were U/S that had to be returned. I'm not talking about the cheapo variety but top of the range stuff. You only have to look at the forum posts here, to see just how common RCDs actually do fail. As i stated they are, or can be, fickle things. Which is why it is always better to have a decent Ze, Ra, Zs on an installation and not rely on RCDs as many are doing these days, especially on TT systems where they become the sole means of earth fault protection...