New wiring - need some help please

[pablo]

Hello!

First of all i am not an electrician but need some help on how to progress with a problem i have. I apologise in advance if i'm not using the right lingo but i hope i'm making everything clear.

We've just had an extension of our lounge completed. Electrical fittings were 6 double plugs on a new ringmain and several spotlights all controlled by 3x 3gang dimmer

Dimmer switches:

Dimmer 1
a) controls a single GU10 50W Spotlight
b) controls 2 GU10 50W spotlights
c) controls a set of 4 50W GU10 spotlights

Dimmer 2
a) controls a set of 4 50W spotlights
b) controls a set of 4 50W spotlights
c) controls 1 GU10 50W spotlights

Dimmer 3
a) controls 1 single GU10 Spotlight
b) controls 1 single GU10 Spotlight
c) controls 2 GU10 50W spotlights

The dimmers are getplc and rated at 60-300W per gang.

Every now and then we hear a loud 'bang' that appears to come from either the ceiling or behind a wall's plasterboard. A millisecond before the bang the lights (everywhere in the
house) dim for a fraction of a second. This could occur once every few hours or even a few times within the space of a few seconds. It also occurs when all lights in the lounge are off and also when there is nothing plugged into the lounge sockets.

The work was carried out by a building company and further inspected by another electrician. He performed wiring tests on all sockets, lights and fusebox and found no problem and he is equally confused. Unfortunately we had no 'bang' while
he was here which may have helped in diagnosis. We have the electrical installation certificate.

The only compatability issue here is the dimmer controls that control only one spotlight at 50W, rather than the minimum of 60W. The electrician said that the problem we have would not be caused by not meeting this threshold; the said light would only flicker.

So we are completely stumped and require opinions assistance from experts!!

Hope you can help.


Paul

 

[LukeScotty]

not come across this before, I sure someone will have an answer for you, check back regularly

regards luke

 

[Girlyspark]

When it 'goes bang', what actually happens at your fuse board - does just one circuit trip (ie lights?) - I'm assuming your lights actually go off, as you haven't actually said that in so many words. Or does an RCD trip? Do the sockets stop working?

The Spark is right about the under consumption of power - the light would just flicker.

 

[pablo]

The lights do not turn off. They just dim for a fraction of a second, immediately followed by the bang and the lights regain full brightness. Nothing happens to the sockets at all - they remain working.

 

[Grae79]

any smells of burning? do you have RCD protection at fuseboard?

 

[pablo]

No smell of burning thankfully!

I have had a look in the garage where the oringinal consumer unit sits and can tell you how the extension is powered:

ORIGINAL WIRING
From meter>Distribution box(Greenbrook 2289 120A 250V Max conductor size 25mm2)>Original fusebox>original house circuitry.

EXTENSION WIRING
From Greenbrook 2289>same thickness cable as above>RCD Volex Protector VB32VB>VB06 tripswitch (extension lighting) and VB32 (extension main).

Seems rather strange that the house lighting dims when this bang occurs if the problem is due to the extension as they are on separate circuits.

Have now turned off the extension lighting and main from the Volex Protector to try and isolate the problem.

 

[TonyM58]

OK Pablo, had a good think about this.

All of the lights dim - this indicates a loss of voltage.

Now, lets suppose you have a damaged cable ANYWHERE in the house. Lets suppose that what you have is an intermiitent short between phase and neutral (due to damaged insulation etc)

When the short 'occurs' the result is that the voltage on your entire system would drop dramatically, as there momentarily the phase and neutral conductors would become 'electrically the same point' and there would be no potential difference (voltage) between them (this is all theoretically)

Now a TV for example, when on, could drop its supply to as low as about 50V without going off. The point i am making is that ALL of your house voltage could be dropping, but it would only be evident on the lighting circuit!

So lets say the fault is on the new wiring, under these circumstances it would still effect the exisiting house circuit.

So why the bang?

If the cables are 'just touching', then at a set point they will get close enough so that current will be able to jump the gap between them. this now a fault current which will rise very rapidly (hence the drop in voltage etc) however as it is a 'weak' path, then the bang is the cables 'blowing' themselves apart again.

you must appreciate this is all hypothetical. And because of the nature of your loads, and the current we are taking about, there will also be an elemant of collapsing magnetic fields (due to basically an inrush of current due to the fault) 'assisting' the spark which is essentially jumping between your cables.

I am trying my best to describe it, but if you ever seen a set of contacts opening on a circuit (like the old contact breaker points in earlier cars), well the spark you see is essentially whats happening in/around one of your cables/connections.

How do you find the problem?

Well if the circuit with the faulty cable is electrically isolated, then the fault wont occur. i.e. with your extension circuits off and you dont get the bang, then i would switch ONE of the cirucits back in (ring main) and observe....... then switch that off and put the lights back on and wait........

Now if you can identify the faulty circuit ( and assuming you do not have test equipment) ( and also assuming you have sufficient electrical knowledge to isolate circuits safely), then I would isolate ,say half the lighting circuit, and wait....

By a process of eleimination you should be able to identify WHICH cable is faulty. Of course, you really need a decent sparks on this - and the original ssparks may not be at fault - its amazing how much damage a nail can do AFTER the sparks has signed the job off

Of course, I could be talking out of my butt (been acused of that before!), but anyone got a better theory??!

Oh, and another thought. If my theory is coorect, it may be evident if you plug a table lamp into a socket and see if that dims at the same time as the house lights

 

[Grae79]

theory sounds...well...sound to me! as u say Tony....bit of testing required

pablo, u said u have the install certificate?? what readings are recorded for insulation resistance?

 

[TonyM58]

Oh and another thought Pablo (been thinking again)

when you got electrical sparks you get localized ionisation of the air surrounding the spark. Essentially this means the air becomes electrically charged, and essenetially 'explodes' as the current iginties it.

I haver seen this with testing high voltage circuits with meters, if there is too much exposed conductor on the meter probe, as the probe gets close enough for the spark to jump, there is what can best be described as ' flash over' with a mini explosion.

If this is happening on your cable ( on a very small scale) it would account for the bang - and I have no doubt were you in a position to observe it, you would see a flash as well!!

 

[Girlyspark]

Great advice from Tony.

Really interesting - keep us posted Pablo - I'd love to know the outcome.

 

[Minky]

Pablo:-
Re the electrical problem you are experiencing,after some consideration,as Tony says voltage drop is the issue that causes your lights to dim.
Not entirely convinced that having all that load on those dimmers is good practice,due to heat build up etc.
I would recommend replacing the dimmers with normal rocker switches temporarily to see if this eliminates the 'noise' experienced.Not convinced that a short is being experienced since there is no mention of you having to reset the overcurrent protective device,when the 'explosion' occurs.
The theory i have with regards to voltage drop may be that your house may be on the same phase supplying another installation with high inrush currents,if you can establish which phase you are on,and compare that to other installations nearby,e.g. neighbours then by enquiry it may be known who is pulling the most load on a particular phase.
It may also help to try to establish a timeline for these events within your installation.

 

[pablo]

OK Pablo, had a good think about this.

All of the lights dim - this indicates a loss of voltage.

Now, lets suppose you have a damaged cable ANYWHERE in the house. Lets suppose that what you have is an intermiitent short between phase and neutral (due to damaged insulation etc)

When the short 'occurs' the result is that the voltage on your entire system would drop dramatically, as there momentarily the phase and neutral conductors would become 'electrically the same point' and there would be no potential difference (voltage) between them (this is all theoretically)

Now a TV for example, when on, could drop its supply to as low as about 50V without going off. The point i am making is that ALL of your house voltage could be dropping, but it would only be evident on the lighting circuit!

So lets say the fault is on the new wiring, under these circumstances it would still effect the exisiting house circuit.

So why the bang?

If the cables are 'just touching', then at a set point they will get close enough so that current will be able to jump the gap between them. this now a fault current which will rise very rapidly (hence the drop in voltage etc) however as it is a 'weak' path, then the bang is the cables 'blowing' themselves apart again.

you must appreciate this is all hypothetical. And because of the nature of your loads, and the current we are taking about, there will also be an elemant of collapsing magnetic fields (due to basically an inrush of current due to the fault) 'assisting' the spark which is essentially jumping between your cables.

I am trying my best to describe it, but if you ever seen a set of contacts opening on a circuit (like the old contact breaker points in earlier cars), well the spark you see is essentially whats happening in/around one of your cables/connections.

How do you find the problem?

Well if the circuit with the faulty cable is electrically isolated, then the fault wont occur. i.e. with your extension circuits off and you dont get the bang, then i would switch ONE of the cirucits back in (ring main) and observe....... then switch that off and put the lights back on and wait........

Now if you can identify the faulty circuit ( and assuming you do not have test equipment) ( and also assuming you have sufficient electrical knowledge to isolate circuits safely), then I would isolate ,say half the lighting circuit, and wait....

By a process of eleimination you should be able to identify WHICH cable is faulty. Of course, you really need a decent sparks on this - and the original ssparks may not be at fault - its amazing how much damage a nail can do AFTER the sparks has signed the job off

Of course, I could be talking out of my butt (been acused of that before!), but anyone got a better theory??!

Oh, and another thought. If my theory is coorect, it may be evident if you plug a table lamp into a socket and see if that dims at the same time as the house lights

Thank you for your time spent with your reply and also for putting it into language that i understand!!

I turned the tripswitches for the extension off to find out whether the noise is actually from the extension wiring and after hearing, what seemed like, a similar noise i got fed up and turned it back on again. There was some existing wiring that was terminated around the area of noise but i safely disconnected that but it made no improvement.

I ought to do this methodically, as you advise, and also be a little more patient!! Tomorrow i shall disconnect the extension once again and take things a little more slowly. At the moment I am sitting upstairs in the computer room with only a lamp turned on to see if mains is as affected as the lighting circuit but sod’s law – no bang!

We even got to the point of thinking that the noise could possibly be due to expansion and contraction of copper heating pipes. Reason being is that a long steel girder had to be placed touching some pipework where there was little other option (due to gap between ceiling downstairs and floor upstairs). But i guess that could be negated as there ‘shouldn't’ be any sort of voltage drop when this happens.

Just to mention, we have had an electrician to do some testing since the lounge was completely finished and he couldn’t find any source of the problem either.

My dad’s an engineer and might be able to get some sort of oscilliscope device to try over a weekend which he said is some kind of grid that you place over a surface and it peaks where the sound is eminating and you can further narrow it down by compressing the grid. I’m just guessing but maybe with that device and one of those electrical cable detectors….maybe a long shot!!

I was wondering whether there is any kind of monitoring equipment that you could connect into a socket or onto a light fitting which has a memory to see if there are any spikes occurring and can record the time. (A lot of guessing going on in this reply, but just thinking aloud!!)

Oh and another thought Pablo (been thinking again)

when you got electrical sparks you get localized ionisation of the air surrounding the spark. Essentially this means the air becomes electrically charged, and essenetially 'explodes' as the current iginties it.

I haver seen this with testing high voltage circuits with meters, if there is too much exposed conductor on the meter probe, as the probe gets close enough for the spark to jump, there is what can best be described as ' flash over' with a mini explosion.

If this is happening on your cable ( on a very small scale) it would account for the bang - and I have no doubt were you in a position to observe it, you would see a flash as well!!

Hehe! I hope it doesn’t come to ripping the walls off as we are quite pleased with the extension itself!! 

@ Grae79:
Thanks for replying. Here is a little link to the relevant page of the installation certificate:

Pablo:-
Re the electrical problem you are experiencing,after some consideration,as Tony says voltage drop is the issue that causes your lights to dim.
Not entirely convinced that having all that load on those dimmers is good practice,due to heat build up etc.
I would recommend replacing the dimmers with normal rocker switches temporarily to see if this eliminates the 'noise' experienced.Not convinced that a short is being experienced since there is no mention of you having to reset the overcurrent protective device,when the 'explosion' occurs.
The theory i have with regards to voltage drop may be that your house may be on the same phase supplying another installation with high inrush currents,if you can establish which phase you are on,and compare that to other installations nearby,e.g. neighbours then by enquiry it may be known who is pulling the most load on a particular phase.
It may also help to try to establish a timeline for these events within your installation.

Thank you for your response too.
Everything works as it should. The only thing we experience is the light dimming very briefly before the ‘bang’ and that’s it. There is no resetting required after the ‘bang.’

You have got me thinking about the dimmers though. I listed the dimmers in the first post as 60-300W but after some research think that they are only 250W rated per gang. With the GU10 50W bulbs, I have been advised to double their wattage when calculating load on each gang on the dimmer. Meaning the sets of 4 spotlights should be treated as 400W load on their particular gang which is over the dimmer spec. I didn’t see this as a major issue as the noise also occurs in the middle of the night when all lounge lights are turned off. However, excusing my ignorance, is there some kind of residual load that exists when they are off?

Also we are on single phase. But again, as this can happen any time of the day, I would have thought it was unlikely to be affected by large neighbouring loads. Will also take your advice on noting times of occurrence.

Thank you all again for your replies. I will post back with results when I have results of isolation ‘testing’ as time and patience allows! Any more thoughts are welcomed. 

Paul

 

[Grae79]

yeah ur dimmers are over and underloaded...want sorted really...TLC have quite a good range...can even have remote controls for them if ur that way inclined. i.e. lazy.

test sheet's not the best but results look a-ok to me...is 'yes' a measurement of impedance? guess polarity's not that important after all huh

 

[TonyM58]

Pablo,

regarding the 'residual load,, then no, once they are off, they are are effectively open circuit, so there should be no current flowing through them, although if my theory regarding an intermittent short circuit is correct, then a high current will flow through it.

Regarding the circuit breakers (MCB's) not tripping. Assuming you have BS EN 60898 circuit breakers 9it will be written on them) and assuming they are type B (so if its a 6A it will have B6 written on it) then they have an instantaneous trip current of 3 to 5 times In (where In is the rating of the circuit breaker)

So even a 6A lighting MCB will take between 18-30A before it will trip instantaneously (as opposed to a sustained overcurrent situation)

Take that same principle for a 32A type B, (your ring main) and your looking at 90-150A before it will trip. the idea of 150A flowing as a 'spark' is horrendous, so it is quite reasonable (all theoretical again) that this could happen without your MCB's tripping.

Regarding monitoring: we used to use 'current clamps' connected to a digital multimeter. the clamp goes around an individual cable (such as the incoming phase to the house from your meter). the current flowing through the cable produces a magnetic feild which is detected by a 'Hall effect' chip in the clamp (basically a semiconductor that is affected by magnetic fields) and converted to a current reading on the meter.

We used to use a meter with a 'hold' function on it. this essentially 'locks' the meter onto the highest reading it recieves.

So....... if you get hold of one (and if your dad's got access to an oscilloscope then he may have access to one), you clamp it around your incoming phase, and just 'act normally' - i.e. carry on use everything as normal. Monitor the meter regularly and you will see the reading change as loads change.

however,....if my theory is correct, then your lights dim, you hear the band , you rush over and check your meter and it should show a sharp rise in current from its previous reading!

Suppose it does, then you start to isolate which circuit the fault is on by clamping it to indivudual circuits from each MCB. Ensure you clamp it to the outgoing phase ONLY, if you clamped it around a flat twin and earth cable, then the magnetic fields around the phase and neutral conductors would cancel each other and you would read zilch!!

Appreciate we are going pretty technical now, but if when the alternative is to start ripping your walls apart.......hope this has helped!


regards

Tony

 

[sivoodoo]

BUMP, I wonder what happened, the thread dies off...
Hope poor pablo got it sorted?