non-RCD circuit tripping RCD

[very_pointy_ears]

Hi,

On top of one of the kitchen wall units we have this arrangement:

the 12vDC output of the power supply then feeds some LED strips under the unit.

This is connected to one of the kitchen light circuits so that the LEDs light up when that particular light is switched on at the wall switch.

This has been fine for over a year, but in the last week, turning the light on has started to randomly trip the RCD in the consumer unit the moment the wall light switch is pressed (it happens approx 20% of the time). The same happens if the lights are already on and I switch the switch in the picture off then on a few times.

Now the weird thing is that this circuit is not on the RCD protected side (single RCD CU).

Wondering whether there was some weird neutral/earth conductivity going on, I tested it with the earth disconnected in the socket - and it still tripped the RCD.

If I unplug it, I can switch the kitchen wall light on and off as many times as I like and the RCD doesn't trip (last night I switched it on/off 50 times in succession without tripping the RCD).

Yesterday I also managed to trip the RCD by plugging in an extension lead that had an amplifier, my laptop PSU and a few other small PSUs plugged into it. This IS on an RCD protected circuit though ... but I can't think of a good reason why plugging this in would trip the RCD per se, given that having that stuff plugged in hasn't tripped it since.

I'm at a bit of a loss here. Any ideas?

 

[very_pointy_ears]

I've just tripped the RCD by plugging my laptop in. I've reset the RCD and typing this on the plugged-in laptop. Something seriously weird is going on.

I have been doing some wiring work, but so far most of that has involved ripping out old stuff rather than installing new.

Maybe tomorrow I'll see if there's a CPC I can disconnect in the CU to stop the tripping, to see if there's a specific circuit causing it - does that make sense?

 

[JohnW2]

Maybe tomorrow I'll see if there's a CPC I can disconnect in the CU to stop the tripping, to see if there's a specific circuit causing it - does that make sense?

No, it doesn't. That would be a potentially very dangerous way of trying to detect the reason for RCD tripping.

Kind Regards, John

 

[crystal ball]

I had a vaguely similar problem some years ago, plugging in various appliances tripped the RCD, the problem was traced to a burnt neutral in the shower

 

[winston1]

So you have a switched FCU feeding a switched 13 amp socket feeding a power supply all off the lighting circuit? What idiot did that? FCUs are not needed on a lighting circuit. 13 amp sockets should not be fitted to lighting circuits (someone will be along soon to say it is allowed, but even so it is bad practice). You don't need 3 fuses in line for some LED strips, nor for that matter 3 switches in line.

Re the RCD trips. There could be a leakage on the RCD circuit not quite enough to cause it to trip until a load is applied elsewhere.

 

[ban-all-sheds]

How about someone coming along to tell you to STFU about your obsession with not putting sockets on lighting circuits?

 

[very_pointy_ears]

After quite a bit of tracking down I've found the culprit to be my UPS. If this is plugged in, then occasionally switching the light on and off trips the RCD. If the UPS is not plugged in, I can switch the light on and off 50 times without the RCD tripping.

Is this a common issue with UPSs? It's an APS SmartUPS 1500, in case that matters.

 

[very_pointy_ears]

So you have a switched FCU feeding a switched 13 amp socket feeding a power supply all off the lighting circuit? What idiot did that? FCUs are not needed on a lighting circuit. 13 amp sockets should not be fitted to lighting circuits (someone will be along soon to say it is allowed, but even so it is bad practice). You don't need 3 fuses in line for some LED strips, nor for that matter 3 switches in line.

What would have been your solution - put a fused outlet in and wire the flex directly into that, leaving a trailing IEC lead ready to plug into the PSU?

13A sockets on lighting circuits - I agree bad as it can overload the circuit, which is what the fuse is for, and the label stating 3A max. Any of those components taken in isolation would be idiotic or pointless, but the combination makes sense, IMO.

 

[winston1]

What would have been your solution - put a fused outlet in and wire the flex directly into that, leaving a trailing IEC lead ready to plug into the PSU?

13A sockets on lighting circuits - I agree bad as it can overload the circuit, which is what the fuse is for, and the label stating 3A max. Any of those components taken in isolation would be idiotic or pointless, but the combination makes sense, IMO.

There is a switch on the kitchen wall, right. So all you need is an outlet plate with a trailing lead. Not a fused outlet as a lighting circuit is already fused or MCB and a 3 amp cartridge fuse will have no discrimination against a 6 amp MCB.

13A sockets on lighting circuits are bad as you agree. A 3 amp fuse in it won't help as explained above. A label does not stop overload, people don't read them.

 

[very_pointy_ears]

There is a switch on the kitchen wall, right. So all you need is an outlet plate with a trailing lead. Not a fused outlet as a lighting circuit is already fused or MCB and a 3 amp cartridge fuse will have no discrimination against a 6 amp MCB.

13A sockets on lighting circuits are bad as you agree. A 3 amp fuse in it won't help as explained above. A label does not stop overload, people don't read them.

Presumably the cartridge fuse might discriminate over a 6A MCB on a (say) constant and long duration 6A current draw?

Having a separate switch up there to enable/disable the feature has been usefully convenient a few times. I agree that there's no need for two switches up there, but it all came down to what was available at the time and it does no harm.

Although unlikely (but probably not significantly more or less likely than just abusing the socket) what's to stop someone plugging said trailing lead into a more demanding appliance? It is a standard power connector after all, with at least a 10A rating.

If we're happy to rely on the 6A MCB to come to the rescue, why all the fuss in the first place? Unless that is that someone gets fed up of the breaker constantly tripping and opens up the CU to replace it with a higher rated one. Where do you draw the line?

 

[EFLImpudence]

I wonder how many sockets the OP has on his socket circuits. Chance of overloading, I'll wager.

 

[securespark]

the problem was traced to a burnt neutral in the shower

Turn the temperature down, then.

 

[very_pointy_ears]

I wonder how many sockets the OP has on his socket circuits. Chance of overloading, I'll wager.

Actually the wiring is **** here - spurs feeding spurs, etc and actually too few sockets to be useful nowadays, making it necessary to use multiple extension blocks.

How many sockets would /you/ limit a 32A ring to?

 

[EFLImpudence]

How many sockets would /you/ limit a 32A ring to?

Using Winston's (flawed) "logic" - two.

 

[very_pointy_ears]

Using Winston's (flawed) "logic" - two.

Ha

How about using regular logic? Things like laptop chargers, LED lamps, DVD players, Sky boxes etc don't draw much power but all need something to plug into.