Buzzing Light No Power - How?

[b1ffb0ff]

Hi

I wonder if anyone has experienced this and may have an answer please!?

In short, when any of the switches disconnects the switched live to the lamps, the lights go off, but a faint buzzing sound can be heard from both lamps – found 30-60Vac between Live and earth – only happened once so far, at night!!!!

What I know:

PME earthling arrangement
The buzzing doesn’t happen all the time, and the first and only time I heard it was last night at around midnight. I am a tall chap and it caught my ear as I walked under the lamp.
Polarity is correct – throughout
Voltage between Switch live and Earth was 60Vac – at time of buzzing
Voltage when it doesn’t buzz is circa 30Vac
Open the circuit breaker with one of the lights switches off – V between SL and N = 1.3Vac
Open the circuit breaker with one of the light switches on – V between SL and N = 0Vac
Two way Intermediate switching works as it should – in all permutations
Insulation resistance is off the scale between P & E and P & N and N & E (installation side)
Measured with digital multi-meter – good one, not tat

Installation:
In my bedroom I have two lights (low energy lamps on pendants – nothing flash) and they are controlled by two number two way switches and one intermediate (MK White plastic single gang switches – the newer ones with smaller switch modules).

One two way is by the door and the remainder are either side of the bed – any switch controls both lights.

The strapping is by way of a 3 core from 2way switch one to Intermediate, and then to the final 2way switch two.

The permanent feed and switch wire are taken from the ceiling rose down to 2way switch one (L1 &L2).

How on earth, when the switches isolate the supply from everything except the lamp and then Neutral, am I getting a potential difference of 60Vac at each ceiling rose between the dead switch live side and earth?

How is it making a circuit when the strapper disconnects the feed to the lighting circuit?

Is it me or is this weird?

Thanks in advance

John

 

[breezer]

if its buzzing its a bad connection if its humming it smells or does not know the words.

humming would imply a transformr

 

[aptsys]

Buzzing from the lamps or the lampholder/rose?

CFLs will buzz a bit if there's some voltage on the switched live as you suggest there might be.

Insulation tests won't be very useful using a DMM, you need to be testing the insulation at 500V.

 

[b1ffb0ff]

It is buzzing from the lamp, both of them infact. I have the same low energy lamps throughout and these were the only ones that buzzed.

I realise that i have not stuck 500v across it, but i would have thought a low insulation resistance, of that magnitude, would have shown up on my multM set to MOhmes - 60v represents a low resistance in my book.

Will borrow my mates tester - wierd how it did it at night though and not day - poor sample i guess need more data - could have been higher supply volts, less demmand?

I was a sparks - still am technicaly - in marketing now though so don't have all the kit i used to have.

TRhanks for the comments.

 

[Spark123]

It will probably be down to capacitive coupling in the cables, this can happen with low energy lamps. If you put a normal lamp in the voltage falls to zero and any others on the same switch should stop buzzing. I'm thinking the only way to go about stopping this from happening is by using the grey or blue conductors as the connection wire to the lamp and the brown/black or red/yellow as the strappers, this would mean the core which connects to the lamp is the opposite side of the CPC. It would be an interesting experiment to carry out, just to see if it would work, something like this:

 

[ebee]

interesting connection in that top switch - the strip connector used.
Not required really.

Sorry back to the topc.

Stray capacitance/ inductance would be my first thought.
If only manufacturers would build a shunt resistor into CFLs we wouldn`t get much buzzing/strobing effects.

 

[breezer]

interesting connection in that top switch - the strip connector used.
Not required really.

there is also one in the intermediate switch, what else would you suggest then?

 

[ebee]

Remove the strip connector and put the 3rd core in C.
The T & E - remoce from strip & C and put the two cores in L1 & L2.
No need for strip connector on the two ways.
Although you still need then on the Int switches

 

[breezer]

Remove the strip connector and put the 3rd core in C.
The T & E - remoce from strip & C and put the two cores in L1 & L2.
No need for strip connector on the two ways.
Although you still need then on the Int switches

but the 3rd core must not touch anything.

if the incoming blue and brown touch (when all switches off) the light will come on

so how can you put anything in C when there is a wire there already?

 

[bernardgreen]

ebee

The object was to keep the CPC between any permanent live wire ( black or brown ) and the switched live ( grey ) so as to reduce the capacitive transfer of energy from any permanent live to the switched live using the CPC as a partial screen.

Putting the live and switched live onto the L1 and L2 means there will be higher capacity between them and hence highe energy transfer into the switched live.

 

[ebee]

Sorry I might not have made my meaning clear.
Remove all connections.
the T & E from the pendant (Feed & Return)
put in L1 & L2.
The three cores put in C, L1 & L2.
At the next (2 way) switch smilarly C, L1 & L2.
Just ensure that C & C are the same core in each switch.
Any intermediates between the two just connect to L1 & L2 and leave the common intact.

 

[breezer]

i am with you, you mean like the last switch, and the feed / returns in L1 & L2

 

[Spark123]

Like this you mean?

The reason I redrew it with the terminal block in the 1st 2 way switch is, as bernardgreen says, to keep the live return to the lamp the opposite side of the CPC to the strappers to try and cancel out any capacitive effects. If wired as above the live return to the light can become next to a strapper depending on the position of the intermediate switch.

 

[ebee]

OK