Voltage readings

[TonyMc]

Went round to a friend's house as they had lost all power to lights when moving a switch. Found out that he had not put the feed into the junction box under the step, so rectified this. However when testing the circuits I got some strange readings and wonder if anyone can shed some light on them

At a light fitting (the only one I managed to test as he is getting the place reboarded and plastered) I got 240v between line and neutral when it was turned on, but still got 65v when the switch was in the off position (same reading with two different switches). I was also getting voltage 110v between line and earth and around 90v between neutral and earth.

I then disconnected the feed at the junction box (that we are assuming goes straight to the consumer unit) and I was getting 240v line to neutral, 100v line to earth and 95v neutral to earth.

The wiring is very poorly done with cpcs not connected in most junction boxes, etc so there is no continuity of CPC.

Insulation resistance test all came out at maximum 999.

It has confused me how insulation resistance results are so good but yet there are strange voltage readings

 

[SPECIAL LOCATION]

The wiring is very poorly done with cpcs not connected in most junction boxes, etc so there is no continuity of CPC.

Insulation resistance test all came out at maximum 999.

It has confused me how insulation resistance results are so good but yet there are strange voltage readings

If you have no continuity of the CPC's..

Then any voltage readings / IR readings are about as much use as a chocolate teapot!

Floating conductors can give all sorts of random voltages as you have no reference to measure against..

And IR to a wire not connected back to electrical earth is not much different to an IR test with both you probes/clips just resting on a wooden desk..

You NEED to back-pedal to basics of doing R1+R2 continuity and polarity throughout the whole circuit before attempting any other test readings..

otherwise you are just pi55ing in the wind.

:coffee

 

[Geoff1946]

What are you testing with?  I would guess a high impedance input multimeter designed for electronics work. These will give all sorts of apparent "voltage " readings on non-connected wires due to capacitive coupling to adjacent live wires.

It sounds to me that your friend should have the wiring renewed before the new plasterwork is done.

 

[TonyMc]

If you have no continuity of the CPC's..

Then any voltage readings / IR readings are about as much use as a chocolate teapot!

Floating conductors can give all sorts of random voltages as you have no reference to measure against..

And IR to a wire not connected back to electrical earth is not much different to an IR test with both you probes/clips just resting on a wooden desk..

You NEED to back-pedal to basics of doing R1+R2 continuity and polarity throughout the whole circuit before attempting any other test readings..

otherwise you are just pi55ing in the wind.

:coffee

Cheers for the reply. I agree I should probably back peddle and do the R1 and R2 but to be honest I assumed I wouldn't get a reading where I tested the light due to seeing the Cpc coiled round the grey sheat in a couple of junction boxes so probably no continuity. The IR was Carried out on the cable I assumed went straight to the con unit (maybe wrong) and as I was getting wiered voltage readings I couldn't understand why the IR was ok. 

What are you testing with?  I would guess a high impedance input multimeter designed for electronics work. These will give all sorts of apparent "voltage " readings on non-connected wires due to capacitive coupling to adjacent live wires.

It sounds to me that your friend should have the wiring renewed before the new plasterwork is done.

Tested with a Megger 1720. Cheers

 

[SPECIAL LOCATION]

Cheers for the reply. I agree I should probably back peddle and do the R1 and R2 but to be honest I assumed I wouldn't get a reading where I tested the light due to seeing the Cpc coiled round the grey sheat in a couple of junction boxes so probably no continuity. The IR was Carried out on the cable I assumed went straight to the con unit (maybe wrong) and as I was getting wiered voltage readings I couldn't understand why the IR was ok. 

Tested with a Megger 1720. Cheers

The golden rule of fault finding... 

When you come across anything strange, unexpected or weird, don't assume anything until you have proven it to be correct with appropriate test equipment and correct test methods.

OR..

to put it another way..

Assume EVERYTHING is either; damaged, broken, faulty, incorrectly connected etc..  

Until you have personally verified it is not broken, damaged, faulty, incorrectly connected etc..

As a side note...

I went to do some work extending a ring circuit quite a while ago..

You would have assumed the circuit was all ok as the customer hadn't reported any problems with sockets not working..

However doing ring continuity tests I found the N & CPC were both broken somewhere around the circuit..

(Customer couldn't understand there was a problem, as from their perspective all sockets worked!)

And further investigations showed that the CPC was in fact broken in two places...

So some sockets basically had no CPC connected back to the CU..

And some of the voltage readings L->CPC were completely random on the sockets with no direct continuity back to the CU..

No good faffing around measuring any voltages...

Had to get back to basics confirming polarity & continuity from the CU all the way around the ring..

So anytime you get thrown or confused..  Get back to basics and assume nothing! 

Guinness

 

[Bruspark]

Which also proves the second rule that even correct voltages in the right places doesn’t prove it’s right .. 

 

[jock565]

assumption in our game kills, verify the continuity of all conductors and then test, if the circuit is not complete then it is not ready to be tested and any results are not worth the time wasted attaining them,......or am I missing something ? btw, how can you do an IR if you dont know where things start and stop ?