RCD tripping when a bulb blows

[John Rumm]

Yes good point...

I would have thought it could be more likely on supplies with poor power factors due to large amounts of harmonic noise (rather than reactive loads).

[Chris Bacon]

yOU THOROUGHLY MODERN THING, YOU. Drat, caps lock. Anyway, advice - that's what Newsgroups are best at!

[dale hammond]

> or if you want to wire and use it as an ordinary light as well:

> Conected to the ligting circuit that powers the lights in the vicinity of

[Andrew Gabriel]

MCB's have an additional rating referred to as their Type. Newer MCB's will be Type B, C, or D, sometimes written as B32, C6, etc when referring to the trip current. Older MCB's will be Type 1, 2, 3 or 4, and types 2-4 very roughly correspond to types B-D.

This Type specifies how much of a surge the MCB will allow though without tripping. Type B is 3-5 times the MCB rating, type C is

5-10 times, and type D is 10-50 times. Type B is most common in residential installations. Type C would be fine on most lighting circuits and in theory might reduce the incidence of MCB trip when lamp fails, but it certainly doesn't stop it completely. Another option would be to fit a 5A BS1361 cartridge fuse carrier in place of the MCB.

[kdband]

Thanks for the explanation.

Tried explaining that the new system is safer but it doesn't stop complaints the next time it trips. I think I'll get hold of a type 'C' and see if that shuts her up.

Thanks John.

[Dave Plowman (News)]

But an MCB can't protect a human.

[Dave Plowman (News)]

Yes. Although I realise a whole house RCD wouldn't protect everyone from death. ;-)

[Smudger]

tony sayer said the following on 23/06/2005 08:01:

Apparently, the reason a lightbulb takes out the MCB when it pops is because the filament will always fail at the weakest (i.e. thinnest point). At the point where the filament fails a spark gap is created between the broken ends of the filament and the voltage gradient is sufficient to ionise the gas in the light bulb. Hence you get an extremely low resistance (and therefore large current) until gravity causes the ends of the filament to fall away. Depending on exactly where the filament breaks, this ionising effect does not always occur, so not every bulb would take out the breaker. Also, the effect is not limited to a particular brand of light bulb - it all depends where the weakest part of the filament is.

Clearly, the current is high enough to trip MCBs on the magnetic part of their action...

...Which takes us onto MCBs in general. MCBs (miniature circuit breakers) are thermal-magnetic circuit breakers and they have two separate tripping mechanisms.

The thermal trip consists of a piece of resistance wire wrapped around a bimetallic strip. As the current rises, the bimetallic strip heats up and starts to move. As the current rises above the rated current of the MCB, the strip will move enough to trip the breaker.

This makes the circuit breaker sensitive to small overcurrents, but is relatively slow to respond to large overcurrents.

The magnetic trip consists of a solenoid, which trips when the current rises very quickly during a short circuit condition. As another poster has said, the rating of the solenoid varies according to whether the MCB is a B-type, C-type, etc.

Thus, the MCB is fail safe in the sense that the two operating elements act independantly.

If the solenoid part fails to operate, the thermal part will trip the MCB eventually.

If the thermal part fails, the resistance wire will more than likely melt and make the MCB permanently open circuit.

If you take an MCB apart, you will find it's quite complicated inside. There is, of course a manual trip as well as the themal and magnetic trips which all have to be integrated into the mechanism. There is also a stack of metal plates, which act as an arc-quencher to protect the contacts in the event of a short circuit.

One last note, if you've got this far. Somebody commented that MCB's don't protect humans the way RCD's do. It's worth noting that the reason we have fuses/MCB's is to protect the cables (not the appliances or people directly). This is an important point. Fuses protect cables and therefore prevent fires. Fires kill humans!

Cheers,

Smudger

[tony sayer]

In article , Smudger writes

Err yes Smudger, it was tripping an RCD I was on about not a MCB. This is an instance where politics and sheer bloody mindedness have resulted in a rather poor install and a BIG issue is being made of bulbs will take the power out every time they blow 'cos they'll trip the RCD!....

[John Rumm]

Or try:

other words go to the modern version of a fuse!

[John Rumm]

It can from indirect contact. Obviously it won't help much for direct contact though.

[tony sayer]

In article , fred writes

Are you suggesting that due to abnormal current flow there is a lag in the current sensing inside the RCD unit?.

CMRR yes, I'm very familiar with this in balanced line working in professional audio work......

[fred]

In article , tony sayer writes

No, just that there is more likely to be an error in the accuracy of the differential current measurement when: a) The current is very large (a short), and b) When the rise time if the current is very fast (again likely for a short) I suppose also there's a c) When current waveform is dirty (rapidly fluctuating - say in a short/arc)

Specs for devices with high CMRR usually have bandwidth or risetime caveats, I've seen it at first hand with opto-isolators where output glitches appeared and the logic fits for transformers too where shunt parasitics destroy the high frequency balance and therefore the common mode performance.

I imagine the most likely lamp condition to result in an RCD trip would be one failing to an arc lasting longer than the trip time of the device.

I am assuming the lamp is on the protected side in this case?

[martinfrompi]

I believe that current thinking is that putting the lights on the RCD side of the unit reduces the risk of electrocution, but increases the risks of a) being unable to get out in a fire; b) falling down stairs in the dark. More people are injured in fires and from falls than from electrocution so the balance of risk seems to be against protecting the lights with an RCD.

Your call of course.

[Dave Plowman (News)]

In other words go to the modern version of a fuse!

I've actually had a rewirable fuse pop when a bulb failed. ;-)

[Dave Plowman (News)]

Yes, I'd heard these arguments. But why would an RCD trip in a fire?

Yup. In practice living in London there's enough waste light around at night to find your way down the stairs here as there are a couple of 'borrowed' lights.

[The Natural Philosopher]

If you have some earth leakage, or worse som RFI filters with live to earth capacitors on them, ort even worse an earth neutral short somewhere.. the voltage surge when a bulb blows CAN indeed trip an RCD.

[martinfrompi]

Fire burns insulation, live shorts to ground, RCD trips.

Of course, if it was the lighting circuit you're b*****ed anyway.

[Owain]

Plastic melts, insulation fails, fault to earth, RCD trip.

Owain

[John Rumm]

Me Too!

Only happened after I had a mains halogen pop though. Never happened with a conventional GLS bulb.

Since I went all MCBs I have had one trip on a conventional GLS failing... the emergency light worked well though ;-)