What happened to my lighting

[MartinLiddle]

Last night everything was fine. Today I had been working in the garden until 5pm and when I came in I tried to switch on a light; irritatingly I can't remember whether it was the utility room or the kitchen, nothing happened so I went and looked at the consumer unit and sure enough a circuit breaker had tripped and it wouldn't reset (it just felt limp). The rooms without lighting were the utility room, the kitchen and the living room. I decided the most likely explanation was a failed bulb so I decided to remove the bulbs in the kitchen and the utility room; they were all LED bulbs and all seemed to be seated properly. I checked the resistance of all 5 bulbs and that just showed an open circuit although a couple showed a reading for a fraction of a second. I then went back to the consumer unit and now the breaker could be reset. The lights in the living room were now working. I started replacing the bulbs in the kitchen and after replacing each one, I turned on the lighting and each time it worked. After replacing the last bulb in the utility room it was still working.

My question is what on earth happened to trigger the circuit breaker and leave it in a state where it wouldn't reset?

 

[Black Hole]

irritatingly I can't remember whether it was the utility room or the kitchen

Hardly matters as they're on the same circuit (apparently).

it wouldn't reset (it just felt limp)

That sounds odd. I don't think it would feel "limp" if the fault condition was still present, trying to reset it would simply refuse to latch.

they were all LED bulbs and all seemed to be seated properly. I checked the resistance of all 5 bulbs and that just showed an open circuit although a couple showed a reading for a fraction of a second.

Well, they would (LED lamp). The brief indication was just the input capacitor charging up.

My question is what on earth happened to trigger the circuit breaker and leave it in a state where it wouldn't reset?

No idea. Perhaps there's a fried mouse in the ceiling void.

 

[gomezz]

That sounds odd. I don't think it would feel "limp" if the fault condition was still present, trying to reset it would simply refuse to latch.

Is that true of all makes and models of circuit breakers?

Just wondering if Martin left out telling us of another step involving turning off the whole consumer unit at some point which is when the circuit breaker then go reset?

As for the initial issue then a nearby lightening strike (had one of those melt a metal power socket box embedded in an outer wall? Or, our old favourite fallback for any mysterious mainframe computer glitch which we absolutely could not explain any other way - a neutrino hit!

 

[Black Hole]

Is that true of all makes and models of circuit breakers?

No idea, just inferring. I might have a different impression of what "limp" means.

Just wondering if Martin left out telling us of another step involving turning off the whole consumer unit at some point which is when the circuit breaker then go reset?

That does not seem likely to me. Individual circuit breakers are independent electromechanical modules fitted on a DIN rail with the supply bus on one side and the outgoing circuit on the other.

 

[MartinLiddle]

Just wondering if Martin left out telling us of another step involving turning off the whole consumer unit at some point which is when the circuit breaker then go reset?

No, it never occurred to me to turn off the whole consumer unit.

As for the initial issue then a nearby lightening strike (had one of those melt a metal power socket box embedded in an outer wall?

No there was no lightening in the area.

 

[MartinLiddle]

No idea, just inferring. I might have a different impression of what "limp" means.

The switch on the relevant breaker could be pushed into the "on" position with zero resistance which is what I was trying to convey by using the word "limp".

My guess is that some time, not too far in the future one of the bulbs will fail.

 

[gomezz]

One of the light sockets experienced some damp which shorted out the fitting? Steam from the kettle perhaps?

 

[MartinLiddle]

One of the light sockets experienced some damp which shorted out the fitting? Steam from the kettle perhaps?

The light fitting is a couple of metres from the kettle and the light fitting and kettle position haven't changed in 20 years.

 

[Black Hole]

there was no lightening in the area

AvP really, but you might want to look up what that that means!

 

[gomezz]

The light fitting is a couple of metres from the kettle and the light fitting and kettle position haven't changed in 20 years.

Presumably the upstairs bathroom plumbing runs close above the kitchen ceiling?

 

[antipodean]

Or, our old favourite fallback for any mysterious mainframe computer glitch which we absolutely could not explain any other way - a neutrino hit!

In a previous job incarnation I was national tech support for the business minicomputer range the company marketed. One particular model used the earliest incarnation of 16k bit dynamic RAM chips, which included a parity bit for each byte. These chips required +12V and +/-5V supplies to operate. The memory was done as 32k byte daughter boards, with each daughter board having a 'parity fail' indicator in the form of a magnetic flip over disc. When a parity error occurred the indicator would be set, and the processor would crash reporting 'Parity Error', so that would be the basis of our engineer being called out. He would replace the relevant memory module and send the faulty one to me. On testing it would come up with a hard error on one bit on the module. If the module was left left on the shelf for around a week and then retested the error would go away.

The hypothesis for this was that the memory cell in these dynamic memory chips held around 16000 electrons in one state, and that an alpha or beta particle (can't remember which of the two) was known to have an equivalent charge, so if one hit the memory cell it dumped charge into the oxide layer that was the insulation of the capacitor that held the current state of that memory cell, which meant it only read one state, and couldn't be changed to the other state required for the other binary value. If the memory board was powered up each day this charge would not leak away, but left unpowered for a weak allowed the charge to dissipate and return the memory chip back to usability. It was also known that the ceramic packaging was known to have low levels of particle radiation, and hence was a possible source of the particle, if it wasn't from somewhere else in the natural world.

The problem went away with the next generation of 16k dynamic RAM chips that required only 5V supplies, and I assume also had a die shrink which made the memory cells too small to be affected by the particles.

So blaming a mysterious glitch on a sub-atomic particle was quite possibly correct.

 

[Climber]

Although I’m not an electrician I have gained some knowledge through experience.
I was an electrical engineer (power).

One rule we always usually followed was first to look at the source of supply and hence, one question I would ask is what type of Circuit Breaker is involved.

• Is it one specifically dealing with an overload.
• Is it a combined one with both overload and residual current device (RCBO)
• Does the C.U. have a separate RCD?

I would find it a little difficult to think that 1) & 2) [overload current] would trip due to a faulty bulb as a significant current would need to flow to trip the c/b.
The other issue concerning 2) & 3) [residual current] is that a bulb usually only has a live and neutral connection, but no earth. An earth cable is usually terminated in the ceiling rose.

An RCD operates when it detects an imbalance in the current flowing in a live and neutral. If there is an imbalance the device will trip with a current difference of 40mA. An imbalance basically determines that if the current is not equal in the L&N then it must be going somewhere else, normally to Earth, but there is no Earth to the bulb.
(An Earth would be connected to any light fitting made of metal though).
An RCD is a device with its main purpose being to protect life.

It could also be a fault somewhere in the circuit cabling. I have come across instances where nails in a wall have punctured a twin and earth cable and caused an intermittent fault (hard to detect).
It could be vermin.

A non-latching c/b can be experienced. I am aware that an overload on a c/b is thermally detected. Bi-metal elements will cause the open circuit (due to heat) and trip the breaker. It may take a period of time for the components to cool enough and reconnect again that will then allow the c/b to be closed. That scenario is the likely reason why the breaker may not latch immediately and point to an overload.
There is the possibility it could be a faulty breaker. I’ve personally not come across any in my line of duty though and that involved using electricians in many instances.

Lastly it could be a transient fault, one that resolves itself without ever knowing what caused it and it may not happen again. There’s always a reason for it though.

 

[MartinLiddle]

Although I’m not an electrician I have gained some knowledge through experience.

Thank you for the detailed response, apologies for the delay in replying but I was out for most of the day yesterday.
[quote}
One rule we always usually followed was first to look at the source of supply and hence, one question I would ask is what type of Circuit Breaker is involved.

Is it one specifically dealing with an overload.
[/quote]
I think so.

Is it a combined one with both overload and residual current device (RCBO)

I don't think so.

Does the C.U. have a separate RCD?

Yes it does.

I would find it a little difficult to think that 1) & 2) [overload current] would trip due to a faulty bulb as a significant current would need to flow to trip the c/b.

I have seen circuit breakers on this consumer unit trip due to bulb failures.

The other issue concerning 2) & 3) [residual current] is that a bulb usually only has a live and neutral connection, but no earth. An earth cable is usually terminated in the ceiling rose.

Agreed that the bulbs don't have an earth.

It could also be a fault somewhere in the circuit cabling. I have come across instances where nails in a wall have punctured a twin and earth cable and caused an intermittent fault (hard to detect).

I have managed to do something like that at other properterty in the past but this house was purchased brand new 20 years ago and has not had much work done to it and i don't think it is likely.

It could be vermin.

I can't rule that possibility out but I have seen no recent evidence of vermin. The only vermin problem I had was in the garage where a mouse had got in and did quite a lot of damage to things that it could chew. Once I was aware of it, it was eliminated quite quickly and I made it a lot more difficult for mice to get in.

A non-latching c/b can be experienced. I am aware that an overload on a c/b is thermally detected. Bi-metal elements will cause the open circuit (due to heat) and trip the breaker. It may take a period of time for the components to cool enough and reconnect again that will then allow the c/b to be closed.[/U] That scenario is the likely reason why the breaker may not latch immediately and point to an overload.

But what would produce an overload and go away when the bulb is removed?

There is the possibility it could be a faulty breaker. I’ve personally not come across any in my line of duty though and that involved using electricians in many instances.

i have come across a single faulty breaker in a different house which would trip very occasionally for no obvious reason. When I felt that I had eliminated all other possibilities, I had the breaker replaced and since have experienced no further problems at the other house.

Lastly it could be a transient fault, one that resolves itself without ever knowing what caused it and it may not happen again. There’s always a reason for it though.

I suspect it will happen again but I have no idea when.

 

[Climber]

Martin - My thoughts regarding a combined O/L and RCD breaker having a trip would be: was it O/L or Residual current that caused the trip. My set up is also a seperate C/B and RCD. and easier to determine the likely fault depending on which one tripped.
I've personally never had a trip due to a bulb blowing, so that's a mystery to me.
Although I said previously that would be unlikely to cause a breaker to trip I would not completely rule it out.

But what would produce an overload and go away when the bulb is removed?

Perhaps a co-incidence as the contacts in the breaker have now cooled down and closed allowing the breaker itself to be closed again.

 

[Black Hole]

I've personally never had a trip due to a bulb blowing, so that's a mystery to me.

That harkens from the days of tungsten. A filament bulb can transient short circuit.

What I don't understand about the "limp" breaker action is not that it won't latch but why it would be limp. There should be a spring biasing the breaker to the off position, with the bi-metal element acting to latch it in the on state.

 

[Climber]

What I don't understand about the "limp" breaker action is not that it won't latch but why it would be limp. There should be a spring biasing the breaker to the off position, with the bi-metal element acting to latch it in the on state.

I actually don't know the answer to that, but when thinking about it there could be some mechanical interlock that is governed when the bi-metalic strip is in a position that breaks the circuit current. Otherwise you could then close the breaker when the contacts are still open. (still warm). That would result in no restoration of supply. So it may just be a feature that will only allow the breaker to close when the bi-metalic strip makes contact. Hence the limp effect?

 

[prpr]

That harkens from the days of tungsten. A filament bulb can transient short circuit.

Yes, I had that every time a bulb blew.

 

[Black Hole]

So it may just be a feature that will only allow the breaker to close when the bi-metalic strip makes contact. Hence the limp effect?

Good idea.