RCD fault

[RF Lighting]

Had a strange one today.

The RCD at my parents house tripped for seemingly no reason. It’s a MEMera 2000 board installed by me about 20 years ago. It’s effectively a split load 16th edition installation, although its done with separate CUs for historical reasons.
The board supplies all the sockets, central heating, shower pump etc. All wiring is PVC/PVC of varying ages.

They reset and it held for about 5 minutes before tripping again.
Reset it a second time and it held for about 90 mins before tripping again, so I went to investigate.

First thing I did was a global IR with the RCD off L&N>E as there were still loads connected at this time. I got a reading of 0.8MΩ. Not ideal, so broke the install down, removed the loads and tested each circuit, and they all tested good.

Put everything back together and the global IR was still down. After a while of trying to work it out, I disconnected the RCD functional earth wire, and the global IR went straight up to 48MΩ, which is probably about for the installation.

I found a new old stock exact replacement in one of my boxes of crap at home - I knew it would come in useful one day! Anyway new RCD installed with functional earth and the global IR remains at 48MΩ

The power has been back on a few hours now and touch wood, everything seems ok. What are the chances that the RCD its self was causing the problem?

Out of interest, I tested the old RCD in situe under load conditions and it tripped at 21mA on a ramp test and the trip times were all within spec.

As I was just arriving at my parents, their across the road neighbour whom I have done work for before came running across saying she needed me as thier RCD had tripped this afternoon, and then again about 90 minutes later, almost exactly the same time as my parents house. What are the chances that this is just a strange coincidence? IIRC i installed her a split load half RCD half RCBO 17th edition board. I’m not sure which device tripped yet, but I said ill bob in and have a look in the next day or so.

Here’s the setup purely for your own amusement.

 

[ericmark]

Seems odd, today on another forum some one saying how there RCD was tripping installed around 1992, and my old house had bouts of tripping then would go years with no trip.

I think the early RCD could trip with spikes on the supply line, often when resetting one RCD the other RCD would trip, and I know in the past I have had a RCD trip, the tester shows as A1 but to show doing something swapped the RCD, tested new one also A1 but tripping stopped, I was looking at my stock of RCD's and MCB yesterday, and thinking I really should throw them all out. House now all RCBO's.

 

[securespark]

I have had problems with RCDs on TT supplies tripping for no apparent reason. I suspected spikes on the supply.

I have also had issues with a Memera 2K RCD tripping oddly. I swapped it out and never heard any more.

Could your folks and their neighbours have a spike issue? If so, why would the issue have disappeared on replacing the unit?

Is your neighbour's board also Memera 2K?

 

[RF Lighting]

I’m not sure what’s been going on. So far no trips since the RCD replacement at my parents house.

Supply is a PILC TNS original from when the house was built in 1958.

The over the road neighbours have a niglon board which was not my choice of brand but these things happen! My parents have spoken to their actual neighbour and she has had no problem at all, although it’s quite feasible that they share a phase with over the road but not with next door. IIRC she has a dual split Hager board, although it’s not an installation I’ve had much involvement with. I’ve not had chance to have a look at over the roads installation yet, but it will be interesting to see what’s going on.

I’d guess that spikes are unlikely but not impossible as it’s a very residential area with no industry any where near.

I’m going to set up the removed RCD on its own at their house and see if it trips. There is a merlin RCBO in the garage on a submains which was unaffected.

It’s certainly a strange one. My only other thought is that they have had trouble with mice previously under the floor in the house. I have previously had on a job an intermittent RCD fault which was caused by rodents contacting a chewed live cable under the floor, but Gary the cat seems to have solved this problem at my parents house as far as we know!

 

[ericmark]

My neighbour was a welder, not a clue if he did welding at home, but it could have well caused the bout of trips, he was an away worker so often not at home.

 

[Harry Bloomfield]

My neighbour was a welder, not a clue if he did welding at home, but it could have well caused the bout of trips, he was an away worker so often not at home.

I have a welder in my workshop, which is protected by RCD - I have never had the RCD trip whilst welding.

 

[ericmark]

I think the 1990's RCD's were far more sensitive to spikes than the modern type.

 

[Harry Bloomfield]

I first became acquainted with them in the late 70's, on an heavy engineering job underground - a massive, dangerous and completely black space without any natural lighting. The entire site, including 110v temporary lighting was fed via RCD, in the Dales, miles from any town. The supply had lots of spikes and as a result the RCD would trip numerous times per day leaving guys in complete blackness. How do you site manager that the risk of an RCD constantly tripping in such circumstances, is much greater than the risk of electrocution?

 

[ericmark]

I first became acquainted with them in the late 70's, on an heavy engineering job underground - a massive, dangerous and completely black space without any natural lighting. The entire site, including 110v temporary lighting was fed via RCD, in the Dales, miles from any town. The supply had lots of spikes and as a result the RCD would trip numerous times per day leaving guys in complete blackness. How do you site manager that the risk of an RCD constantly tripping in such circumstances, is much greater than the risk of electrocution?

In the tunnel in Hong Kong I was working on, not carrying your miners lamps was instant dismissal. One never relied on the TBM lighting or supply. Not that a RCD would really work very well with an IT supply.

On Sizewell 'B' I made an error, 110 volt temporary lighting my boss had decided better using fluorescent lights. So each one was 60 watt, so approx 0.5 amp, did not want to be right on the edge, so seemed reasonable to has 25 lamps on each supply, should have been between 13 and 14 amp.

All installed switched on, and they looked good, then hour latter report stopped working, found MCB off, so thought some one messing around turned back on all seemed OK, tripped again, so this time took the clamp on ammeter, 21 amp they were drawing, had to change all the tappings from 110 to 127 volt in the lamps, still took 15 amp, but it is so easy to make an error.

However 1970's not sure about RCD, I think from memory the ELCB-v came out around 1960 and the ELCB-c around 1980 may have this wrong, but even when we went from v to c most were 100 mA S type, the 30 mA RCD came a lot latter. I returned to UK around 1990 and there were still loads of ELCB-v in use, but they were being replaced with the ELCB-c, where I worked we often had a 5 amp and 1 minute feeding a number of 1 amp at 30 seconds which in turn feed 100 mA type S which in turn feed 30 mA at 40 mS, but a nail in the cable would trip the lot. Sizewell 'B' was first time I had seen so many RCD's, I had seen them back in 70's in Theatre Clwyd so yes were around, but went to work in Algeria and Falklands from 1980 to 1990 and on return they were everywhere.

 

[securespark]

Gary the cat

Fantastic name for a cat!

I was going to suggest a shared phase between your folks and the people over the road.

Eric......someone at the Wylex factory in Sharston told me once the first Wylex RCD (the big old wide grey thing) was first released in 1974, but not sure if that is accurate.

 

[Adam_151]

I'd always disconnect the functional earth when doing the global IR, most makes of RCBO it drags it right down, tbh I don't know with these, and these are the only 61008 devcie I know of with the functional lead.

Its odd that the new one does not drag it down to 0.8 as the figrues look right... 48meg for the install and 0.8meg if you forget to disconnect the functional earth. Could yuou have reversed the polarity of the IR tester between these two tests, perhaps there is a diode in the functional earth path of the RCD and it only drags it down to 0.8meg on one polarity... should be easy to check, can you do the tests on the removed RCD that you have on your bench?

Think the MEM RCDs of that age did have poor EMI filtering, I have heard of a caretakers PMR radio taking a load of RCBOS when he was stood near the board