The IET is carrying out some important updates between 17-30 April and all of our websites will be view only. For more information, read this Announcement

This discussion is locked.
You cannot post a reply to this discussion. If you have a question start a new discussion

intermittent RCD tripping advice please

Hello there everyone,

I wonder if anyone would be kind enough to advise on the best course of action with a fault that my father has.

He’s had an intermittent RCD tripping fault since late last summer.

The RCD regularly trips anything from 2 hours to 4-5 days. There doesn’t appear to be any obvious reason for the trips - appliances, weather, time of day etc.

Installation is twelve years old and was all newly installed when their house was renovated. The CU is a Europa and has two rows of 10 slots, each with RCCB protection. The bottom row is the one that keeps tripping - top is fine.

They have had a few things added fairly recently like AC and a new microwave. The microwave seems the closest to the timeline of problems starting.

They do have a Whirlpool bath that needed a new control lead (low voltage logic control stuff), due to water rusting the male/female connector, but that has now been replaced - no water ingress to the main PCB control box. However, Dad found that when turning the bath supply back on that it would sometimes trip straight away, but normally it stays on when idle. I became suspicious of the bath when I learned of the rusted control cable and tripping and advised him to fully isolate the bath to see if the tripping stopped. Isolating the bath does seemed to have reduced the frequency of the tripping, but not stopped it (maybe it’s just adding a tad more leakage, which just pushes the RCD over the limit?). There are a few boxes under the bath (PCB control box, LED control box, air blower/heater). Note: the bath wasn’t connected when the latest IR test was performed. However, I’ve seen the power trip just by turning on their outside lights, for example.

First electrician came a few weeks ago and did some very basic testing. He was on his way home from another job and didn’t have much time. He tested the RCD, said it was fine, but replaced it anyway (Clipsal) to rule it out. He told Dad at the time that it was a N-E leak and that it could be anything causing it. He tried for a while to get him back, but he was too busy with an industrial install.

A different electrician visited this week and spent two hours there looking at the install. I had the chance to speak with him this morning. He said that the IR test didn’t look good on three of the circuits, which are all the ring mains. He didn’t test leakage. He tested the outdoor sockets which were all fine. Dad thinks the meter said 360 for one of the ring mains, but no idea if this is correct - take that with a pinch of salt! He also discovered that top RCD was stuck, but freed it and said it was fine (it hasn't tripped since, but the bottom has twice).

He recommended that Dad worked one room at a time, unplugging everything, leaving it for a week and using this approach to try and locate the problem. Dad isn’t keen on this as he’s a bit older and not the most technical person! However, I’m happy to spend some time there to do some testing, if it will help.

Given the randomness of the tripping, I asked him if he thought it might be caused by cumulative leakage across all the ring mains or a single faulty appliance. He felt it was a possibility.

When Dad wasn’t keen on the testing option, he suggested upgrading the existing CU with RCBOs. Cost is ~£700. As I believed would be the case, he confirmed that doing this won’t necessarily fix the problem, but it would potentially pinpoint the circuit - although, at this point, surely we are suspecting the ring mains, or in particular the two ring mains on the bottom CU row that trips. He told me that it is not cost effective for him to diagnose the fault, but he felt there was value in somebody like me doing the testing and was supportive of that idea.

Would it be worthwhile getting a clamp meter and testing all of the appliances on the ring mains - starting with everything connected to the sockets on the two ring mains of the bottom CU row that is tripping? I understand that I’ll need to adapt an extension lead so I can clamp just the L-N.

What about clamping the L-N at the CU, taking the background readings of the leakage? Could you leave the clamp on and then go around and unplug things one at a time, looking for a large drop and potential source of the problem(s)?

I can understand that you might have a faulty appliance on one ring main, but all three? Likewise, how likely would it be that all three ring main/circuit cables are damaged in some way?

Appreciate any insights, thoughts or advice on this problem.

Many thanks in advance,
Richard

Parents
  • Thank you both very much for your replies.

    What do you think of this meter?

    https://www.amecal.com/product-page/mini-ac-leakage-current-tester-st-9810

    It has 100µA (0.1A) resolution, which I believe should be good enough for this task? The jaws aren't huge, but maybe that's a good thing for the scope of the testing here? Should I get it calibrated?

    I will be going down there after half term to do the testing, so I will report back once I'm there.

    My plan will be this:

    - Test the whirlpool bath

    - Test the heating pump, washing machines, kettle and outside lights

    - Test all the appliances on the three ring mains, but focus on the "bottom two" tripping side first

    For outside soffit lighting (e.g. LED down lighters) or anything that doesn't have any earth, should these devices be tested?

    Thanks again,
    Richard

  • 100uA is 0.1mA, not 0.1A, (!) but that is the last flickering digit of the fruit machine, not the accuracy. You need to be able to see changes at the 10mA level, I 'd expect that sort of thing to do it.

    I've not used one of those model but I have something like it by Dilog that was really useful for about the first  year, and is now on the 'repair me' pile, as something seems to have failed internally.

    Unless it is avery cheap extra I'd be minded not bother with the cal cert personally, as you are really interested in trends of rising or falling as things are added or removed, not so much the absolute numbers.  I have a short extension lead I use where the outer jacket of the cable is removed to expose the brown blue grellow cores for a 6 inches or so  for just this sort of electron hunt for suspect appliances.

    Not having a CPC does not mean that there is never another  path to earth - if you put L and N thorough the core side by side you will (to a pretty good approximation), see the difference, much as the RCD does, and if what comes back does not equal what went out, then the difference  has presumably got on or off on the way round ... TVs can lose current down the antenna cables, computers used to lose current down the phone lines before WiFi, and all sorts of funniness.

    Be aware that you can have current flowing in your earthed services that is not from your own L-N loop - there is another post today about diverted neutrals I suggest you read the 1st PDF if yo are not aleady up to speed with the idea. Combined with an NE fault this can make your RCD trip or not based on the neighbours loads.

    If this is the case, then reading currents in the grelow wire may not indicate a fault that an RCD would see, but the L-N test will.

    Mike.

Reply
  • 100uA is 0.1mA, not 0.1A, (!) but that is the last flickering digit of the fruit machine, not the accuracy. You need to be able to see changes at the 10mA level, I 'd expect that sort of thing to do it.

    I've not used one of those model but I have something like it by Dilog that was really useful for about the first  year, and is now on the 'repair me' pile, as something seems to have failed internally.

    Unless it is avery cheap extra I'd be minded not bother with the cal cert personally, as you are really interested in trends of rising or falling as things are added or removed, not so much the absolute numbers.  I have a short extension lead I use where the outer jacket of the cable is removed to expose the brown blue grellow cores for a 6 inches or so  for just this sort of electron hunt for suspect appliances.

    Not having a CPC does not mean that there is never another  path to earth - if you put L and N thorough the core side by side you will (to a pretty good approximation), see the difference, much as the RCD does, and if what comes back does not equal what went out, then the difference  has presumably got on or off on the way round ... TVs can lose current down the antenna cables, computers used to lose current down the phone lines before WiFi, and all sorts of funniness.

    Be aware that you can have current flowing in your earthed services that is not from your own L-N loop - there is another post today about diverted neutrals I suggest you read the 1st PDF if yo are not aleady up to speed with the idea. Combined with an NE fault this can make your RCD trip or not based on the neighbours loads.

    If this is the case, then reading currents in the grelow wire may not indicate a fault that an RCD would see, but the L-N test will.

    Mike.

Children
  • Hi Mike

    Yes, sorry, I did mean mA not A! User error.

    Ah, the old 'repair pile'... I wouldn't hold out too much hope for the meter lasting many years, but if it helps us find the fault then it will have paid for itself.

    I figured the cal cert would probably be for certification etc., so that will help to save on the cost - thanks for that.

    Will be using the extension lead solution as you describe for the appliances.

    OK, I really wasn't sure about a missing CPC - makes sense though about an alternative path. I knew about the TVs (have also seen some posts on this), but, ah yes, I'm old enough to remember modems and I can see how that could happen too!

    I'll check out the diverted neutrals post too.

    Thanks again
    Richard