Voltage between earth and switched line terminal on one way light switch

S

Sting Ray

• Jul 24, 2014

• #1

After installing some LED fittings in a bathroom and disabling an electric shower circuit I carried out a check on the whole installation. All fine until I used my Di-LOG probe to probe the switch plate screws to check that I had not pinched any live wires in the light switch back boxes around the house (not sure what made me do that?). One switch lit up the voltage present light so off it came. No trapped wires but using my Megger MFT1730 I see around 20 volts between the earthed back box and the switched line when the switch is open, 244V when closed. Disconnected the cable from the ceiling rose and the switch and carried out insulation tests between all conductors, everything >999M. Re-connected and the 20 volts was back.

Does anyone know what I'm seeing/experienced this before and is it real? Or should I just not be so inquisitive!

 

telectrix

-

Mentor

Nearly Esteemed

Arms

Joined

Sep 5, 2010

Messages

73,534

Reaction score

71,830

Location

cheshire/staffordshire

• Jul 24, 2014

• #2

probably induced voltage. try using an analogue voltmeter. it'll probably disappear. was the switch in question 2 way?

 

OP

O

Octopus

• Jul 24, 2014

• #3

Turn off the other breakers and it'll probably dissappear!

 

OP

S

Sting Ray

• Jul 24, 2014

• #4

One way and all other breakers were off too. Induced from where?

 

telectrix

-

Mentor

Nearly Esteemed

Arms

Joined

Sep 5, 2010

Messages

73,534

Reaction score

71,830

Location

cheshire/staffordshire

• Jul 24, 2014

• #5

any cables running in proximity of your switch cable. as i said before. try an analogue meter. the lower input impedance will drain away any stray voltage.you may see an initial "kick" of the needle should then settle at 0V".

 

OP

S

Sting Ray

• Jul 24, 2014

• #6

Will try analogue as you suggest

 

HandySparks

-

Mentor

Arms

Joined

Jul 22, 2011

Messages

4,333

Reaction score

2,656

Location

Hampshire

• Jul 24, 2014

• #7

Get yourself one of these. No chance of any phantom voltage lighting this one up.

 

OP

G

Guest111

• Jul 24, 2014

• #8

Definately sounds like induced to me,thats the problem with using DMM,s many years ago I had a fault on a car and was convinced it was the fuel pump,the fluke meter said 12v going to pump but pump not running,anyway the AA man who came out wasn't convinced and tested using a bulb on 2 bits of wire,guess what no power.Turned out it was a duff relay under the dash.DMM's have their place but sometimes you just need that extra bit of load.

 

Lucien Nunes

-

Mentor

Esteemed

Arms

Joined

Jul 2, 2013

Messages

6,778

Reaction score

14,809

Location

London for the time being

• Jul 24, 2014

• #9

Am I being dopey or did the OP say that back box is supposed to be earthed? If it is, then induced voltage(from PL to SL) would explain a reading with the switch off, as the LED driver wouldn't be guaranteed to present a near-zero forward voltage L-N like a filament lamp. But it doesn't explain the original presence of voltage on the screws, which points to a disconnected CPC. There could still be enough leakage for the MFT to read 240V L-CPC with the switch on.

What was the Zs at the switch?

 

OP

G

Guest111

• Jul 24, 2014

• #10

Lucien Nunes said:

Am I being dopey or did the OP say that back box is supposed to be earthed? If it is, then induced voltage(from PL to SL) would explain a reading with the switch off, as the LED driver wouldn't be guaranteed to present a near-zero forward voltage L-N like a filament lamp. But it doesn't explain the original presence of voltage on the screws, which points to a disconnected CPC. There could still be enough leakage for the MFT to read 240V L-CPC with the switch on.

What was the Zs at the switch?

Click to expand...

Depends,he said he'd initially used a DiLog test lamp,now if its the same one I had they do automatic continuity test,so maybe it was the continuity light that was on,reading through neutral to earth? Just another possibility to throw into the mix.

 

Lucien Nunes

-

Mentor

Esteemed

Arms

Joined

Jul 2, 2013

Messages

6,778

Reaction score

14,809

Location

London for the time being

• Jul 24, 2014

• #11

But he did that first test to the screw with the switch in place, so he would not have had access to anything but the CPC, which should never light an indicator. We need to know what that test actually was.

 

OP

G

Guest111

• Jul 24, 2014

• #12

Lucien Nunes said:

But he did that first test to the screw with the switch in place, so he would not have had access to anything but the CPC, which should never light an indicator. We need to know what that test actually was.

Click to expand...

very true,but will we ever find out?

 

OP

S

Sting Ray

• Jul 25, 2014

• #13

Zs at the switch and the outlet is 0.61, Ze 0.2, giving an approximate cable length of 13.5m which is about right as this room is directly above the main board

 

OP

S

Sting Ray

• Jul 25, 2014

• #14

I know I can buy this via the web but thought I'd try some of my local distributors (Edmundsons, Senate, CEF) yesterday. No one had heard of this item let alone stocked it although they were all keen to show me fine collections of hen's teeth and rocking horse ****! "What's an analogue voltmeter?" one said LOL.

 

OP

S

Sting Ray

• Jul 25, 2014

• #15

Incidentally, whilst re-testing the switch again yesterday with the DiLOG, sometimes the voltage indicator (forked lightening symbol) didn't light up but would always do so if I touched any wall. Temperature yesterday was 31C! and all the walls in the house are of dry lining construction.