Live - Neutral reverse from plug in tester

So, for some reason the C.P.C. at the sockets is 55 Volts above N. The earth is deficient in some way. That accounts for the 185 Volts from L to E rather than the expected 240 Volts. Curiouser and curiouser.

Z.

Good, so N to local earth (plumbing rads etc )is 0v, and just the circuit CPC is flapping about at about quarter of the way up the L-N voltage. My money is on an open CPC, and a voltage divider formed by the capacitive coupling from L-CPC and N-CPC.

In the middle of a ring this means 2 breaks in the CPC, one on either side, or something really silly, like the back box is earthed but not the socket.

I agree with Mike, it looks like an open circuit cpc (earth) at that socket. Is it possible the terminal screw has caught the sleeving rather than the conductor? 

Regards,

Alan. 

At least no local metal work like rads are reading this 55vlts to a N   point. and all oultes have 240 L-N voltage so no under volts present for appliances. I know telephone line operate arounf 50 volts and IEE regs recognise that 50 volts or under are not considred danger to life. 

If it is as most of us suspect a broken c.p.c. with capacitive coupling to other conductors providing the voltage, then be aware that the values displayed on the voltmeter can be extremely variable. One meter might show 55V another 15V and another 115V - all quite validly. The problem is that the capacitance is normally of very low values - pico Farads - so the current that can flow is tiny, thus the internal impedance of the voltmeter can significantly alter the circuit being measured. Even the best voltmeters draw a little current from the circuit and if there is only a tiny current available the voltage will reduce even before the meter gets a chance to measure it - it's all down to the details of innards of the voltmeter - which can vary considerably.

   - Andy. 

Agreed, most likely broken CPCs with a phantom voltage being measured with a multimeter.

Some voltmeters put extremely low loads on voltage tests which can be a plus on some electronics circuits.

However for testing such as we do it can be misleading.

A loading of a few watts might make the measured voltage start to collapse.

There can be quite a marked difference when using voltmeters with a 10M (ohms) input impedance/resistance and  the older 10/20 K per volt or 1 2 10 K per volt input resistance.

The old Drummond test lamps with actual filament lamps as indicators usually did a pretty good job of collapsing these “phantom” voltages. 

Stray capacitance (or inductance, or insulation resistance) can sometimes play havoc with bog standard multimeters.

Alan Capon: 
 

I agree with Mike, it looks like an open circuit cpc (earth) at that socket. Is it possible the terminal screw has caught the sleeving rather than the conductor? 

Regards,

Alan. 

If the CPC sleeving has been caught twice in two different fittings it would effectively disconnect the  CPC between the fittings.

yes, as a first bash, for typical 2,5mm PVC cable, expect about 100pF of capacitance core to core per metre length (so you'd need hundreds or maybe thousands of km of it to make a motor start capacitor- it is really a very small effect !!) The core to core capacitance  does vary a bit between makers, not being a well controlled quantity, more of an accident of core geometry and insulation thickness, but the microamps that flow at 50Hz are enough to register on a modern digital meter (10 or 20 megaohm input loading typical ) and also to cause some LED fittings and some CFL lamps to pulse in a ghostly way when supposedly off. Confusing if you are not aware, especially on 2 way lighting, with different adjacent cores live or connected to the lamps in the two ‘off’ states. 

A resistor or a capacitor of perhaps 0.1uF or so is sometimes useful to provide a load to collapse these picked up voltages, on short lengths sometimes the pick up can also be modulated a  bit with contact to conductive fingers, but it is a brave soul who does that, as the degree of voltage collapse is not guaranteed.

Mike.

The answer is to drop the circuit out of the consumer unit and test it starting with end to end continuity tests and go from there.

Fault finding with a plug in tester and a multimeter is not impossible, but it’s a bit of a nightmare.

ebee: 
 

. . . The old Drummond test lamps with actual filament lamps as indicators usually did a pretty good job of collapsing these “phantom” voltages. . .

The new LED version of the Drummond test lamp, developed for the Electricity Supply Industry, the MTL20, has two buttons, one on each probe. Pressing the buttons together increases the current drawn for this reason, and is excellent at collapsing phantom voltages. 

Regards,

Alan.