When bulb blows, earth leakage circuit breaker is tripped

[John Rumm]

Its called a fuse! ;-)

[<me9]

This used to happen /many/ years ago with fuses. Then they came along with bulbs that had a built in fuse to prevent this happening. Do incandescent lamps no longer incorporate a fuse, or is it just the halogen ones?

[Guy King]

The message from contains these words:

Well, the incandescent bulbs[1] we use here appear to have fuses in and it still happens - so do you have another explanation? I suspect the fuse still doesn't react as fast as the MCB.

[1]I think we've got two left - apart from those in the cooker hood. Got several live spares in the cupboard though.

[Ian Stirling]

You can get a pack of 10 lithium coin cell powered LED torches, from ebay, for a couple of quid.

Scattering these round the house so that you've always got one to hand if the power goes out can help.

Personally - as part of the rewiring effort, I'm putting in a single LED uplighter per room. (single 5mm LED) These go on when the power goes off, for whatever reason.

[David Hansen]

On Wed, 13 Dec 2006 12:47:21 GMT someone who may be Guy King wrote this:-

Do you? Why?

A generalisation is that a fuse reacts much faster than an MCB during a high current fault, but an MCB reacts faster than a fuse during a low current fault (an overload).

[Guy King]

The message from David Hansen contains these words:

Because when the lamp blows it still takes out the MCB even when there's an internal fuse. For all your cunning reasoning, real life experience suggests that having a fuse in the bulb doesn't prevent the breaker breaking.

[David Hansen]

On Wed, 13 Dec 2006 18:49:50 GMT someone who may be Guy King wrote this:-

As others have said, fuses (in consumer units) can also be "taken out" in such circumstances.

I made no comment on fuses in bulbs. The comparison is between fuses in consumer units and MCBs in consumer units.

[Guy King]

The message from David Hansen contains these words:

Sorry, I thought you were saying that MCBs no longer pop when a light bulb blows because bulbs now have fuses.

I was actually answering "B Thumbs" who said in reply to me saying that blowing bulbs take out MCBs "This used to happen /many/ years ago with fuses. Then they came along with bulbs that had a built in fuse to prevent this happening. Do incandescent lamps no longer incorporate a fuse, or is it just the halogen ones?"

So my reply /was/ about bulbs with fuses though it may not have been aimed at you.

[<me9]

I have seen (and heard) a 100A service fuse blow in preference to a 6a mcb. (The (large) substation transformer was about a fathom away.)

[Andrew Gabriel]

Hum, It's sufficiently more complex than that I don't think any such generalisation can be valid (and if you look at the trip curves, _in general_ I would say it's the other way around). For example, what to you mean by "react"? Actually detect the error condition, or finish clearing the error condition, because those two are quite different between MCBs and fuses, and even quite different between different types of fuses.

Lamp base fuses seem mostly to be of the bare wire type, which are going to be similar response to a BS3036 semi-enclosed (a.k.a rewirable) fuse. Occasionally, ballotini fuses are used, which have different error detection and clearing times, more like that of a cartridge fuse.

Once an open wire fuse melts, it can take a number of mains cycles before the vapour has dispersed sufficiently for the arc to be quenched. Meanwhile, an MCB can operate in less than a half cycle, whilst the fuse arc is still running. If you look at a lamp which has tripped an MCB, you will see its fuses are blown, but they were unable to quench the arc before the MCB tripped. It's a case of a current which grossly exceeds both the fault current ratings of both protective devices, and they both trip. I've never seen a lamp with fuses trip an MCB without also blowing its lamp base fuses (both of them if is has two).

[David Hansen]

On 14 Dec 2006 08:45:34 GMT someone who may be [e-mail] (Andrew Gabriel) wrote this:-

I remain happy with that generalisation.

At the risk of starting a, "mine is bigger then yours", discussion, not something I am keen on. There is a saying, if you think you understand something then try writing computer software to model it. If the software works then you really do understand the subject. I have written and maintained computer software to model electrical devices, though only up to 33kV.

I didn't once mention a rewirable fuse.

[Steve Firth]

That's bollocks. I've written several computer models, all of them modeling with varying degrees of accuracy systems of assorted complexities. I can't pretend to have understood the subject well for every model.

And no, I wasn't coding someone else's models.

[Andy Wade]

The principles here are fairly straightforward, it's obtaining and applying the relevant data that's the problem.

At a particular fault current:

(i) if the operating I^2*t of the MCB is less than the pre-arcing I^2*t of the fuse then the MCB will discriminate in favour of the fuse (i.e. breaker trips, fuse does not blow);

(ii) if the operating I^2*t of the MCB exceeds the total clearing I^2*t of the fuse (pre-arcing and arcing) then the fuse will discriminate in favour of the MCB (fuse blows, breaker does not trip);

(iii) if the operating I^2*t of the MCB exceeds the pre-arcing I^2*t of the fuse, but does not exceed its total clearing I^2*t then both devices will operate (no discrimination, fuse element melts but the arc is extinguished by the MCB).

Now all you need is the relevant I^2*t curves and knowledge of the fault level. I^2*t data for MCBs should be fairly readily available from the manufacturers, but for lamp fuses... (I've never seen any published).

[Ian Stirling]

Not to mention that unless you actually do understand the model, and have verified it against underlying facts, you can come up with a model that seems to work, but doesn't except in the test cases.

Modelling something doesn't help, if your understanding is flawed, and you are just modelling your understanding, if your understanding of the system is self-consistent.

[Ian White]

To add a bit more detail...

I^2*t (pron: "I-squared-T") is a measure of the electrical energy that is let through the fuse or MCB during a current surge. Some of that energy is absorbed by the device itself, either to heat the wire in a fuse, or to give an electromechanical kick to the moving parts in an MCB.

In a fuse, the pre-arcing I^2*t represents a "point of no return", where enough heat has been absorbed to melt the wire, so it is definitely destined to break. However, it takes a while longer for the molten metal to flow, and for surface tension to pull the ends apart.

Thus you can easily have a situation like Andy's case (iii) where the current surge takes the fuse beyond its point of no return while also operating the MCB. Then, as Andy says, it is often the MCB that breaks the circuit, before the fuse has time to pull apart... but the fuse is still found to be melted.

[David Hansen]

On Thu, 14 Dec 2006 23:13:50 +0000 someone who may be Steve Firth wrote this:-

Ah, proof by assertion and rudeness in three words. Excellent.

Noted.

[David Hansen]

On Thu, 14 Dec 2006 23:59:30 +0000 someone who may be Andy Wade wrote this:-

It is.

Fuses built into lamps are not something I have discussed, not the least because I have never seen data on them.

[Tony Williams]

In article , Andrew Gabriel wrote: [snip]

I was going to do a post earlier on in this thread to say that we've never had a blown bulb trip the MCB. That was, until last night, when it happened with an almost new 100W bulb (about 10 days old).

I've just knocked it apart. As you say, destruction is complete. The support wires are there, but the filament has disappeared, as have both fusewires in the stem of the bulb.

SWMBO did note that it didn't do the normal 'plink' as it went, it did a nasty-sounding 'splat'. That could have been the arcing.

[Ian_m]

Two solutions :-

1. Change MCB (I am sure you meant MCB not ELCB) to type C. as this has a higher over-load/time trip. This worked for me in my hall lights (4 x 60W). Got fed up of coming home, turn on hall lights and flash and still in dark. In fact very dark as the IR triggered outside light is on same circuit, so is the garage lights which makes getting to the breaker very hard.

1. Change to low enrgy bulbs.

[Steve Firth]

Life must be so difficult for you, blubbering every time someone says boo!