MCB fault or dodgy circuit??

[ban-all-sheds]

if you do decide to replace the MCB, i would suggest you get some cheap, insulated screwdrivers.

I'd suggest getting some good quality, VDE certified ones.

Jaymack, i know its a bit off the original topic now, but why do you say it isnt safe? I dont have copies of the hse docs. The meter I have is CATIII rated, double insulated, 600v. It is marketed at electrical testing.

I've used it for testing the lighting circuits (other occasions) and found it very reassuring to know for sure the circuits are off. Far safer than common DIY alternative which is using a volt stick or neon screwdriver(!).

Cheers.

Not to say it isn't safer to use a DMM than a volt stick (yuk) or a neon screwdriver (even worse, should be banned), a DMM is still not a recognized tool for testing dead. Recognized testers such as RS 410-0097 conform with the british standard and are permitted for testing dead.

I think the DMM will be fine. If it's verified to be working immediately before, and when turning the main switch off causes all the lights to go out, all the appliances to stop working, and the DMM shows the bus-bar as dead, I think it's safe to assume that it really is dead.

Since you said the MCB is 5A I assume it's a plug-in replacement for a rewireable fuse.

70s/80s vintage, possibly - so it could easily be a Type 1, which would have been 5A, not 6A.

Solderer - Howard is right that you need to be cautious, so turn off the CU main switch, carefully remove the cover and see what's what.
Or beforehand tell us what the CU is, and/or post a photo, someone may well be able to reassure you about what you'll find inside.

 

[mapj1]

The regulations on meters for work use on mains circuits require to meet tighter standards, including things like not having enough metal exposed at the tips to be remotely useful, and internal fusing rated at thousands of aps safe rupturing capacity.
As you are presumably doing this DIY, I can assure you the meter is not intrinsically dangerous, but you could not employ someone and ask them to use it. remember that legally a company cannot waive liability. A private individual can however decide to do so, but only for himself.
However if you do use your DMM for mains testing, do be very, very, careful as careless use could indeed make it potentially dangerous, more so than a "proper" one, if you put it on ohms or amps accross the mains or something equally daft, and a short accross the mains is not pretty - you can be burnt or even blinded when showered with hot metal. Usually it just goes flash bang, but surviving it once does not make it a good idea to try again.
I'd try and arrange a loop of wire out from the CU so you can actually measure the lighting circuit resistance or ideally the on-load current. Remember also most tungsten filament lamps have take about 10 times the running current when cold (as the absolute temperature is 290K when cold and perhaps ~3000K when white hot and resistance is roughly linear with absolute temperature above the Debye point this should not surprise you). Warm up is about 1/10 to 1/5 second depending on wattage.
Circuits with dimmers, which can keep the filament much colder than normal indefinitely, have very high peak currents for short duration, some faster trips do not like this.
For these reasons on larger circuits, or where many lamps are switched together, a breaker with a high surge rating is often required, it need not indicate a fault, but a measurement of circuit resistance or load current will tell you what is happening more surely.

 

[solderer]

Thanks Mike. This is a really interesting point about the dimmers and lamp temps as I do have a set of 5 x 40W bulbs on a dimmer on that circuit, and also another 100w on a dimmer. This could well be a contributary factor.

However, the aspect which is a pain is that it wont reset easily rather than it trips in the 1st place, so perhaps i'll still find there is a problem with the MCB or the circuit.

The circuit hasnt tripped again since i wrote initially, so i'm still alive, having not been tempted to tamper with the CU

I'll write in again if i seriously consider changing the MCB. I can tell this is more dangeerous than the usual household wiring tasks - due to the proximity to the live feed and lack of isolation - so maybe i would fork out for someone qualified to take the risk ...

 

[Adam_151]

As long as you turn off the main switch, and keep away from where the tails come into it, and providing the tails arn't badly terminatated leaving lots of exposed live copper, then you will be ok, you can also have a look near your meter to see whether there is a separate isolating switch, but these arn't all that common.

 

[Jaymack]

Remember also most tungsten filament lamps have take about 10 times the running current when cold (as the absolute temperature is 290K when cold and perhaps ~3000K when white hot and resistance is roughly linear with absolute temperature above the Debye point this should not surprise you). Warm up is about 1/10 to 1/5 second depending on wattage.

I assume that you are saying here, that tungsten has a negative temperature co-efficient of resistance.

 

[HDRW]

Remember also most tungsten filament lamps have take about 10 times the running current when cold (as the absolute temperature is 290K when cold and perhaps ~3000K when white hot and resistance is roughly linear with absolute temperature above the Debye point this should not surprise you). Warm up is about 1/10 to 1/5 second depending on wattage.

I assume that you are saying here, that tungsten has a negative temperature co-efficient of resistance.

No, a positive coefficient - as it gets hotter, the resistance increases. So the resistance is lower at switch-on which is why bulbs most often blow when they are turned on.

If the resistance fell as it got hotter, you'd get thermal runaway and the bulb would blow immediately, making incandescent lights impractical.

Cheers,

Howard

 

[mapj1]

All metals have a positive resistance coefficient, as it heats up the atoms start to jiggle about with thermal energy, and the electrons find it harder to get accross. Think of trying to get accross the London Underground - is it easier when everyone else is sitting still, or if they are all milling about? - exactly.

Above a certain temperature (called the Debye temperature) the resitance becomes dominated by this effect, and becomes a simple linear function of temperature, until the material melts when the atoms jiggle so much they come unstuck from each other.

A few semiconductors, start with very few free electons to carry current, are poor conductors when cold, but the heat loosens up some electrons that were trapped at lower temperatures to carry more current. These materials have a negative coefficient. However, that is the exceptional behaviour, and such things have to be specially made by delicately balancing the chemistry.

Hope that I haven't upset too many experts with the "Noddy does solid state physics" lesson!

 

[ban-all-sheds]

All metals have a positive resistance coefficient

I think there's a U/Nb/Zr alloy which has a negative one.

 

[Adam_151]

Odd. I belived the same as mapj, will probably try and google that, ban.

On a similar but different note, have you seen the video of the conducting glass demonstration thats supposedly floating around the 'net?

Adam

 

[ban-all-sheds]

No, but I did see some transparent (well, OK, vaguely translucent) steel on Tomorrow's World some years ago...

 

[solderer]

Hey, Physicists,

there is one thing though, when the dimmer is initially turned on low, obviously the dimmer circuit is keeping the voltage low (by the thyristor only turning on in the early part of the cycle), and so i would assume the current is still pretty low despite the bulb being cold.

Maybe the dimmer circuit ends up presenting a badly behaved load for the MCB and this doesnt help?

Anyway, i'm still illuminated

 

[mapj1]

Very true, but you still need a beefier triac than you might think, perhaps a 20A one for a kilowatt of load. Note that the filament temperature is not a simple function of power dissipation either - in a vacuum lamp, such as some low voltage aircraft lamps are, the power raditaed off goes as fourth power of abs Temp , but in anything with a gas fill (most ordinary light bulbs), convection gets going inside the envelope after a few seconds, making the P vs T roll off rather faster, and in quartz halogen where the deposit and dissolve reaction is on-going, it can be a very time dependant thing.
So the practical upshot is that at half time power (i.e. waveform sliced at the crest of the wave) Is MUCH dimmer than percieved half brightness, and the filament is actually typically at 800-1200 degrees, so current is then around 3 times what you might expect at full temperature.
Hence noddy rule of thumb of take RMS current and quadruple to find triac currect rating if operation at dim setting is expected for any length of time. There is more to it than meets the eye, as many a dissapointed home made dimmer maker will tell you.

 

[ban-all-sheds]

To say nothing of how some cheap'n'nasty 12V power supplies behave when presented with a waveform chopped in the wrong place.

OOI - does anybody make a dimmer that puts out a synthesised sine-wave of lower amplitude rather than a chopped one?

 

[HDRW]

OOI - does anybody make a dimmer that puts out a synthesised sine-wave of lower amplitude rather than a chopped one?

You mean a Variac? Not synthesised, obviously, but it does precisely this job.

I think a variable-output pure-sinewave invertor would be ridiculously expensive, and far too big to be useful in a domestic environment.

Cheers,

Howard

 

[ban-all-sheds]

You mean a Variac? Not synthesised, obviously, but it does precisely this job.

I think a variable-output pure-sinewave invertor would be ridiculously expensive, and far too big to be useful in a domestic environment.

Unlike a variac, of course....