RCD Trips

[wiggles]

Hi,

I've had one RCD trip yesterday and 2 tonight. Of course, this also coincides with my alarm battery being flat, so everytime it trips, the alarm goes off. As you would expect, it's only tripping at night! I'll be off in search of a new battery tomorrow.

Everything has been fine and dandy for years, and this has only occured since I changed an MCB out on the RCD protected side of the CU from a 16 to a 20 AMP. This my be coincidence but could a lose wire or similar be causing this?

The first trip occured when I switched the dish washer on at the issolation switch. After that trip, I switched it on and off several times and ran it without any issues.

Tonight, the next trip occured when I switched the tumble dryer off at the issolation switch. I then flicked the switch back on in panic. I reset the RCD. About 10 minutes later, the RCD tripped again so before resetting it, I issolated all devices in that area (tumble dryer, dish washer and washing machine).

In daylight I'm goign to check that all the MCB's and wires are tight in the CU. Other than that, can anyone advice on anything else I can check or if I should get a pro in?

Thanks.

 

[Newboy]

and the obvious first question - why did you change the MCB.

Please don't say because the old kept tripping!

 

[wiggles]

I should have explained that. The MCB that I changed was for a radial circuit in the garage which should have always been a 20 amp anyway. It's wired in 2.5mm It was originally a 16 amp. I changed it for a 20 amp as the initial surge from my new compressor was tripping it out. The compressor has not been plugged in during any of the trips.

You've also reminded me that I added a 16 AMP socket onto that existing garage radial. I had intially discounted it as it's switched out from the circuit via a DP switch, but perhaps a wire on the supply side has been pinched behind the switch. I'll check that also.

 

[DaftPunk]

It could be a duff connection. Check the new MCB you fitted has actually clamped onto the busbar. It's easy to install an MCB with both parts of the clamp in front of the busbar. The screw will tighten up and feel like you've made a good connection when really you've missed the busbar altogether and left a high resistance connection.

 

[bernardgreen]

Could be a Neutral to Earth fault on another circuit, Outside lamp full of rain water. A heavy load on any other circuit on that neutral bar will create current through the fault which will lead to the RCD tripping.

Neutral_Earth_trip

• bernardgreen
• 16 Jan 2011
• 48

 

[ericmark]

It could be no fault at all. I have had two RCD's protecting the house since 1992 and every round and again I have a spate of trips without there being any real fault. Spikes on the supply can cause this and those spikes may be generated from outside your house as with a thunder storm.

The toaster is one culprit where a bit of old bread can give an earth neutral fault. Frost free freezers are very hard to identify faults with the de-frost heaters as switched on at a random time.

So former ensure items are unplugged when not in use. Latter swap supply to other ring even if temporary with extension lead and see if fault swaps to other RCD.

As to changing a MCB from 16 to 20 amp it is not only the cable size which matters but also the loop impedance which should be tested to see if with in range before changing. This will depend on distance to garage.

Likely 19 meters is the limit with a 20 amp MCB loop impedance line - neutral around the 0.82Ω assuming coming in at 0.35Ω and the line - earth impedance no more than 2.3Ω assuming using B type and 1.15Ω using C type and 0.57Ω with a D type.

It's not as simple as just looking at cable size and what the run is like. You do need to measure first.

 

[JohnW2]

Likely 19 meters is the limit with a 20 amp MCB loop impedance line - neutral around the 0.82Ω assuming coming in at 0.35Ω and the line - earth impedance no more than 2.3Ω assuming using B type and 1.15Ω using C type and 0.57Ω with a D type.

Do I take it that you mean about 109 metres (with 2.5mm² on a B20 and assumed Ze of 0.35&#937 ??

Kind Regards, John

 

[wiggles]

Thanks for the replies guys. The CU is located in the garage, so the radial in the garage (the one I changed to 20 amps) is very small. It will be under 5m in total length.

There has been nothing conntected in the garage during the trips other than the gas boiler.

I checked all the MCB and connections in the CU earleir and they all appear to be fine.

We have a farily new fridge < month old which is the frost free type. We've had various issues with it since we bought it, so nothing would surprise me.

So far today we have been able to run the dishwasher and tumble dryer together without issues. RCD tripped when the washing machine was turned on, so will try again when the dishwasher has finished.

 

[JohnW2]

So far today we have been able to run the dishwasher and tumble dryer together without issues. RCD tripped when the washing machine was turned on, so will try again when the dishwasher has finished.

If you are getting RCD trips in response to switching on various different appliances, bernard may well be right in suggesting that the problem could be a neutral-earth fault on one circuit (any circuit - not necessarily those supplying the appliances in question) protected by that RCD. If so, and if the trips persist, you will probably need an electrician to find the cause of the fault. Are there any sockets, junction boxes, fused connection units, lights etc. anywhere (e.g. outdoors, maybe even in the garage) which might have suffered from water ingress?

Kind Regards, John

 

[wiggles]

Thanks John.

Washing machine is currently running on it's own without any trip issues.

There are 3 outside lights but all are mounted on house walls and are connected to the non-RCD side of the CU.

I have a single outdoor MK weatherproof socket. It's connected to the garage radial on the RCD side and isolated by a DP switch. It's currently off and was yesterday as well.

My CU is a split load and looking inside it's a bit of a mess. The last electrician has fitted 2 MCB's which do not fit properly. The one I bought was an MK which seems to be a much better fit. Am I correct in thinking I would need a whole new CU to get RCBO's?

 

[JohnW2]

Thanks John. ... There are 3 outside lights but all are mounted on house walls and are connected to the non-RCD side of the CU.

If they are on the non-RCD side then, give or take very esoteric possibilities, they would not be the problem.

I have a single outdoor MK weatherproof socket. It's connected to the garage radial on the RCD side and isolated by a DP switch. It's currently off and was yesterday as well.

The DP switch would obviously stop that affecting anything.

My CU is a split load and looking inside it's a bit of a mess. The last electrician has fitted 2 MCB's which do not fit properly. The one I bought was an MK which seems to be a much better fit. Am I correct in thinking I would need a whole new CU to get RCBO's?

Not necessarily. What make (and, if you know it, model) is the CU - and/or could you post of photo of it?

Kind Regards, John

 

[wiggles]

Hmm. I'm rather suspecting the fridge.

I just looked at it 10 mins ago and according to the display the freezer is -18 and the fridge is 4 degrees which is strange as all the food in the freezer had defrosted and the internals of the fridge are warm.

I powered the fridge off and back on again and the display now shows 12 and 9. Earlier at about 11AM before the leccy tripped it was making a strange noise. I don't think the inverter is running any more.

I guess if there are no more trips, that accounts for the problem.

 

[ericmark]

Likely 19 meters is the limit with a 20 amp MCB loop impedance line - neutral around the 0.82&#937; assuming coming in at 0.35&#937; and the line - earth impedance no more than 2.3&#937; assuming using B type and 1.15&#937; using C type and 0.57&#937; with a D type.

Do I take it that you mean about 109 metres (with 2.5mm² on a B20 and assumed Ze of 0.35&#937 ??

Kind Regards, John

No I worked out the volt drop for a radial at 20 amp not a ring so yes do mean 19 meters.

As to freezer problem it does seem there is a problem. I have also trusted the display on my inverter freezer and fridge/freezer although I did put a sensor in the freezer and measure I was actually testing the sensor rather than the freezer before taking to my mothers to test her freezer.

It is rather worrying that the display in your case is wrong. I note mine has a warranty for 10 years on inverter and motor but rest has only a two year warranty. Mine was faulty as new and the manufacturers engineer had to attend to fix it. Seeing how much he had to dismantle to swap a control board can't see a repair being cheap.

For my mothers the freezer cost £110 and to swap the thermostat cheapest quote was £45 plus parts. I ordered a new thermostat at £5 very easy to fit and tested it with the temperature controller I use for brewing beer.

As to testing I used a plug in energy meter. With the inverter there is very little inrush at start. So 53W running motor and 215W when de-frost heater cut in so one could see with easy if the freezer had defrosted from energy meter readings. With the old fridge/freezer the start current was 2259W so unless you happened to be watching when de-frost started you didn't know if it had or had not done a de-frost cycle but with new inverter it's easy to see de-frost cycle has taken place.

 

[JohnW2]

Likely 19 meters is the limit with a 20 amp MCB loop impedance line - neutral around the 0.82&#937; assuming coming in at 0.35&#937; and the line - earth impedance no more than 2.3&#937; assuming using B type and 1.15&#937; using C type and 0.57&#937; with a D type.

Do I take it that you mean about 109 metres (with 2.5mm² on a B20 and assumed Ze of 0.35&#937 ??

No I worked out the volt drop for a radial at 20 amp not a ring so yes do mean 19 meters.

Oh, I see. Since you didn't mention VD, but only EFLI, I assumed that's what you were talking about. I didn't actually previously actually do a calculation, but 19 metres sounded so small (EFLI-wise) that I guessed it was probably a missing digit' typo. .. but, yes, if the circuit supplies lighting, and if you want to stick to the guidance limit of 3% VD (which we have both recently agreed is essentially daft), then one would end up with a calculated maximum length of about 19 metres (or about 31 metres for the 5% without lighting). Having now done the sum, EFLI-wise one could have about 83 metres - and EFLI is (IMO) what 'really matters', safety-wise.

Whatever, this is obviously all moot, now that we know the cable is only about 5 metres long!

Kind Regards, John

 

[wiggles]

No trips since the fridge died, so it looks fairly certain now. I have to say it's been a very expensive piece of junk at just under 1k bought about a month ago. We've had problems with it since day 1 with ice production being very slow. Water line freezing. Stuff in the fridge freezing. Motor being loud. Samsung engineer has been out twice to make adjustments to it and told us that they are having a lot of problems with that model. It's all culminated in a lost of wasted food also.

For me though, the last few days of trips have been the worst as for some reason during the power failure my home server decided to corrupt itself and I've spent nearly a day trying every trick I can think of to restore a backup which turns out to be corrupted also! Total pain and waste of time.

Anyhow, back to the subject. Here's a picture of the CU. On the internals it has an earth bar all the way along the top. A neutral bar to the left and right and a split bus bar at the bottom.

Can't see a make on it. It was a generic screwfix model from many years ago when they first started online.

The 'blue' MCB's are the original ones. The two Wylex ones were added about 5 years ago by an electrician. They don't fit very well and one of them actually popped off the bus bar when I fitted the replacement MK MCB.

What's really annoyed me is that when looking in the CU, the electrician has just cut the wires of two old circuits and left them hanging in the box instead of removing the wire even though they both lead to the kitchen where he was working at the time.

View media item 87632