3-way lighting has residual voltage of 50v when switched off

[helps]

I have added an extension to my home and now have 3-way switching in the hall. I have followed the manufacturer's wiring instructions and they are the same as those wiring diagrams shown here on the forum and on other sites.
I will have an electrician come replace the existing consumer unit and approve the installation prior to signing off for building warrant, but it shouldn't be that complicated as the following is true-

The 3-way lighting is measuring 238-240V when any of the hall lighting is switched on, but when it's switched off there is still a residual voltage of 42-50V and this causes the LED downlighters to remain dimly illuminated.

All the switches allow the hall lighting to be turned on or off, but there is a residual voltage. There are 3-downlighters and I have fitted one while the other two cables are left disconnected. I fitted one down light just to check it was working ok before covering up with plasterboard.

The new extension has new harmonised cabling installed on all circuits including the lighting.

The lighting wiring checks out as the other 5 rooms have multiple lighting points and some with 2-way lighting and there are no similar issues.

I've read a few of the forums but haven't found the answer yet.

Does anyone know what the solution would be?

 

[endecotp]

Are you measuring the 50V with the LEDs in-circuit, or with them removed?

 

[securespark]

How are the strappers wired?

 

[ericmark]

This is quite normal. We should use matched transmission lines to ensure no power is lost, but we don't so capacitance and inductance in the connecting cables for the frequency used with not be matched as a result there will be some small amount of power in the cables even when switched off.

With standard bulbs this was not a problem but with LE|D bulbs taking so little current it can cause the light to come on dim.

I have seen in previous posts on the subject where matching capacitors are used to combat this problem fitted in the switches but I have not seen how the size is calculated to form a correctly terminated system.

Transmission lines is a complicated subject I did do it in university but my son has my books now.

It would seem some LED bulbs and CFL have circuits built in the deal with this problem and when I swapped LED's in my own two way lighting the fault went away.

It is two way switching even if you have 20 switches to control the same light each switch only has 2 ways. I only say that to help looking for diagrams etc.

I seem to remember from the last post with same problem there are capacitors designed for the job.

 

[ban-all-sheds]

I've read a few of the forums but haven't found the answer yet.

//search.diynot.com/forum_sear...tbytime=0&author=&search=Search+Forum&stype=0

 

[plugwash]

This is quite normal. We should use matched transmission lines to ensure no power is lost,

Matched transmission lines don't "ensure no power is lost".on the contary using matched transmission lines tends to greatly increase power consumption (which is why noone even thinks of doing it in power systems).

Transmission line theory is totally unnessacery in the formulation of either the problem or the soloution. The wavelength of 50Hz with a wave propogation speed of 10^8 m/s is 2x10^6 meters so on the scale of a building (or frankly even a city) we are quite safe treating a mains power cable as a single capacitor and a single inductor.

inductance is negligable in this scenario since very little current is flowing and the frequency is low.

The affect we care about is capacitance. Capacitance between the wires can act in paralell with the switches allowing a tiny current to flow when switched off. With most loads this has no noticable impact but some electronic loads with non-ohmic characteristics can have problems.

I have seen in previous posts on the subject where matching capacitors are used to combat this problem fitted in the switches but I have not seen how the size is calculated to form a correctly terminated system.

You put the capacitor across the light not the switch. You don't need to make a "correctly terminated system". you just need to make sure the voltage developed across the light is kept sufficiently low. That means that the capacitance you place across the light needs to be much greater than the capacitance in the cable.

Unfortunately it seems capacitance figures for mains cable are hard to come by. http://www.moeller.net/binary/ver_techpapers/ver949en.pdf quotes a value of 0.3uF per KM for "2 core control cables".

Fortunately we can afford some overkill here.

If we assume that the capacitance in our cable is as above, that the cable will be 100m long and we want to reduce the voltage at the fitting by a factor of 100 then we would need a 3uF capacitor.

 

[bernardgreen]

It is advisable to have a resistor in series with the capacitor to limit the current into the capacitor when the switch operates anywhere in the mains cycle other than at the zero volt point in the sine wave. This reduces wera on the switch contacts.

The resistor if selected correctly also acts as a fuse should the capacity fail to become a short circuit across the mains.

 

[JohnW2]

Transmission line theory is totally unnessacery in the formulation of either the problem or the soloution.

Quite so - not that many people reading this will be familiar with such theory, anyway! At the frequency involved, I'm not even sure that those running the grid need to think too much about transmission line theory! As you say, all that matters in the context of this discussion is capacitance.

Kind Regards, John

 

[plugwash]

Transmission line theory is totally unnessacery in the formulation of either the problem or the soloution.

Quite so - not that many people reading this will be familiar with such theory, anyway! At the frequency involved, I'm not even sure that those running the grid need to think too much about transmission line theory!

AIUI things get "interesting" when you start talking about continent-sized grids.

 

[bernardgreen]

With the advent of electronics that provide efficient DC to AC conversion at the Megawatt level more and more long distance power transmissions are using DC to avoid the need to phase synchronise the entire network.

 

[JohnW2]

At the frequency involved, I'm not even sure that those running the grid need to think too much about transmission line theory!

AIUI things get "interesting" when you start talking about continent-sized grids.

Well, yes, I can believe that - continents can be pretty big! However, I think that essentially underlines the point we were both making about the irrelevance of transmission line theory to 50Hz except when the distances involved are truly enormous!

Kind Regards, John

 

[JohnW2]

With the advent of electronics that provide efficient DC to AC conversion at the Megawatt level more and more long distance power transmissions are using DC to avoid the need to phase synchronise the entire network.

That makes sense.

Kind Regards, John