Mains electric shock from plastic light switch - no fault found by electrician - what next?

[jimster99]

Hi all! This is my first post and I have registered to ask for some help with troubleshooting an electric shock I got from a plastic light switch outside a bathroom today.

I turned the switch off with my left hand and got what felt like a mains electric shock through the switch into my left arm. It had the distinctive "throbbing" feel of mains electricity and I felt it mainly through my left arm and left torso. I pulled my arm away and my rib cage still feels a little sore this evening. It was definitely not a static shock. The apartment has an RCD box which didn't trip.

The light switch itself is plastic but has metal screws (although I don't think I touched them). I was wearing slightly damp socks on a thin carpeted floor and my hand was possibly a bit damp from washing my hands in the toilet so my resistance was probably relatively low.

An electrician has now been to visit and said he couldn't find anything wrong and he seemed to suggest I couldn't have had a shock. He very happily touched the switch and it was fine. The wires in the switch appeared to be screwed in properly and the same in the light. The switch is mounted in a metal frame but he couldn't find any live voltage where there shouldn't be with his two-pronged multi meter. He did suggest that a more thorough check of the property and this is being scheduled.

This may be unrelated but in the morning before the shock my wife had noticed that the light in the bathroom controlled by the switch (which is a ceiling mounted multi-LED type) was pulsing on and off approximately once a second when switched off. When turned back on it was working fine but then still pulsing when switched off. The light then started working normally again so we thought it had cleared up.

Any ideas as to what can possibly have caused this? I didn't imagine it but with the electrician not finding anything wrong, I am flummoxed! Note I am handy with a multimeter but only a hobbyist and I normally steer well clear of working on mains electricity.

Pics of the light switch are below:

 

[snowhead]

There doesn't appear to be a Neutral in the switch box so did the electrician test from the switch screw to another known earth, if not how did he /she establish the metal switch box was not live, other than touching it?

And did he / she investigate whether there's a nail or screw through a live cable in the floor, which you may have made contact with and with earth on the switch screw?

The current required to trip an RCD is at a "painfull level" so there may not have been enough current passing through you to trip it.

 

[littlespark]

Not saying this is what has happened, but sometimes the top and bottom lug of the backbox can touch the middle terminal of the screw and make the box, and the fixing screws live.... Taking the switch off has stopped them touching. The lugs can be bent back, or broken off.

This does raise the question of why the RCD didnt trip, and the electrician should have checked its operation. Although, looking at the age of the switches... the lights may not actually be on the RCD.

It may be that the earth wire in the lighting circuit isn't connected to earth anywhere, and this really needs investigated. Hopefully a more in depth check will find out.

A photograph of the circuit breakers in your consumer unit might help there.

 

[LukeD]

I hate it when i see bend back lugs top/bottom. I wont ever fit them and given the chance will tear them out. And this also answers why the screw covers can be a good thing !

 

[snowhead]

Is that a metal trim on the floor across the doorway and were you standing on that?

 

[Dave OCD]

There's also a slim chance it could have been a small arc escaping from between the switch and plastic surround as that switch would quite possibly be around 30 years old.

 

[DPG]

There's also a slim chance it could have been a small arc escaping from between the switch and plastic surround as that switch would quite possibly be around 30 years old.

That's my thinking as well. The older switches aren't as resistant to being used by wet hands.

 

[Lucien Nunes]

the light in the bathroom controlled by the switch (which is a ceiling mounted multi-LED type) was pulsing on and off approximately once a second when switched off. When turned back on it was working fine but then still pulsing when switched off.

This seems highly significant. Some lights do flash annoyingly when off because they are hyper-sensitive to cable leakage but I assume yours doesn't normally do this. It strongly suggests electrical leakage across the switch for which a likely cause, being outside a bathroom, would be water entering the switch and forming a partially conductive path between the contacts. By the time you touched the switch, the moisture path causing the leakage through to the light seems to have been driven off, but a conductive path might still have existed between at least one contact and the exposed surface of the switch rocker.

As to why the RCD didn't trip during a shock to earth, I can only suggest that if the RCD tested good (did the electrician carry out tests on it?) then the current was probably below the trip threshold. You can still get quite a nasty belt at 20mA, unlikely to affect your heart but it does hurt. The current would have been limited both by the resistance of the moisture track in the switch and by that of the damp-but-not-wet socks / carpet path between you and earth.

Had I been the electrician, I would have immediately carried out a 500V insulation test between all permutations of switch contacts in an attempt to locate any remaining sign of leakage paths within. MK switches of that era can easily be taken apart and reassembled so I would also have inspected internally with the same objective. Obviously as time went on, if moisture was the culprit, all evidence would disappear without trace.

 

[jimster99]

A photograph of the circuit breakers in your consumer unit might help there.

Thanks for the reply! A photo of the (ancient) RCD box is below. I can't test it until tomorrow (wife is working late on office VPN and has forbidden me from disconnecting her...) but I will check it out as it's clearly an old unit and has not been tested by me before - although it has tripped a few times in the past.

Is that a metal trim on the floor across the doorway and were you standing on that?

It is indeed a metal trim! Well spotted and perhaps my foot was on it while I touched the switch (not certain) but if so, it could be the answer as to what my feet were conducting into! I will get my multimeter out and do some (careful) tests tomorrow on that strip. Is it worth covering it in something like gaffa tape as a temporary precaution or is that not going to help?

The RCB:

 

[jimster99]

There doesn't appear to be a Neutral in the switch box so did the electrician test from the switch screw to another known earth, if not how did he /she establish the metal switch box was not live, other than touching it?

The guy was in a bit of a hurry (emergency call out at 5pm probably at the end of a busy day) so he only used a simple multi-meter inside the switch box and also tested by putting his hand in (this made me slightly nervous...). I'm not sure how (if) he checked to another known earth but perhaps he just didn't.

His testing unit was a basic looking multimeter device with two fixed prongs on the end and a series of lights showing live voltage at different marked levels. It beeped to show continuity when he poked it into the switch box (not sure exactly what continuity he was checking) and none of the voltage lights lit up when he touched it onto the switch box. He didn't do any other tests that I could see.

And did he / she investigate whether there's a nail or screw through a live cable in the floor, which you may have made contact with and with earth on the switch screw?

No he didn't check that but it's a good suggestion and I'll have a check to make sure nothing is peeking through. His only comment was that because it was carpet it was impossible for electricity to travel through it. Someone else has suggested perhaps my foot was touching the metal strip between the bathroom and hall carpets - that also sounds like a possibility.

This seems highly significant. Some lights do flash annoyingly when off because they are hyper-sensitive to cable leakage but I assume yours doesn't normally do this. It strongly suggests electrical leakage across the switch for which a likely cause, being outside a bathroom, would be water entering the switch and forming a partially conductive path between the contacts. By the time you touched the switch, the moisture path causing the leakage through to the light seems to have been driven off, but a conductive path might still have existed between at least one contact and the exposed surface of the switch rocker.

As to why the RCD didn't trip during a shock to earth, I can only suggest that if the RCD tested good (did the electrician carry out tests on it?) then the current was probably below the trip threshold. You can still get quite a nasty belt at 20mA, unlikely to affect your heart but it does hurt. The current would have been limited both by the resistance of the moisture track in the switch and by that of the damp-but-not-wet socks / carpet path between you and earth.

Had I been the electrician, I would have immediately carried out a 500V insulation test between all permutations of switch contacts in an attempt to locate any remaining sign of leakage paths within. MK switches of that era can easily be taken apart and reassembled so I would also have inspected internally with the same objective. Obviously as time went on, if moisture was the culprit, all evidence would disappear without trace.

Thanks - this is really helpful. The light does not flash normally. It was so unusual that I filmed it. The RCD wasn't tested but I will check it myself tomorrow. There was no 500v insulation test carried out - I will see if this is possible.

I'm nervous about leaving this unsolved for any period of time. If it is moisture getting inside the switch, what's a good approach to prevent this in future - replace the switch? Or get a full thorough check up carried out?

 

[Lucien Nunes]

By all means check the RCD with its test button, but I would have expected the electrician to have run a series of tests using his test instrument to analyse its performance. To me that is a no-brainer after a shock-to-earth that did not cause a trip.

 

[littlespark]

Sounds like he used a two prong voltage tester, rather than a multimeter…
He still should have had his proper test kit.

It’s a lot of circuits through one RCD, and over time, it can get a little sensitive.
You may want to consider a new consumer unit which will separate the circuits to one RCBO each…. But you’re looking at high hundreds of pounds.

A new switch to eliminate moisture creep, and bend back the top and bottom lugs on the box. Check earth continuity back to board.

And dry your hands before using the switch.

 

[DPG]

I would definitely get the switched changed for a more modern equivalent. After the appropriate tests have been carried out.

And as Littlespark says, try not to use rocker switches with wet hands.

 

[Pretty Mouth]

I have a customer who had a similar problem. Occasionally, perhaps once or twice a year, the lights controlled by the problem switch would strobe brightly when off, disco style. This would last for maybe a couple of hours, during which time he claimed that touching the switch would give little shocks, which seemed to be from the switch rather than the (metal) switch plate. Eventually the problem would disappear without trace, and the shocks would stop.

They had previously had another electrician come and look at it, and he couldn't find anything wrong. When the problem happened again, the customer called me. By the time I arrived at their house a few hours later, the problem had disappeared again, but he showed me footage on his phone of the strobing.

I tested everything I could think of that could be affecting the lights, but could find nothing wrong. External loop impedances were fine, Zs fine, earthing to the metal switch plates was fine, no loose connections etc, IR for the circuit was fine, IR between switch terminals and the plate was fine, RCD trip times as expected. A total mystery.

Couple of notes: the switch, at the bottom of the stairs, was 2 gang, both being 2 way with other switches. Both sets of lights were affected by the strobing. Property about 20yrs old.

 

[OnlQQker]

Just curious about the led light. Is it in a loft ceiling or is there another floor above this?
If it is in a loft do you have tanks up there?
Did the electrician look inside the light?




Any chance of a picture with the bulbs OP?