Odd voltages

[gazdkw82]

Sorted a shorting fault at a pendant earlier this evening.

Whilst i was there they asked if I could look at the wall lights that have never worked since they moved in 2 years ago.

2 wall lights. Getting 244v line to earth, 0v neutral to cpc but 74v line to neutral.

The 74v line to neutral is my guess as to why any lamp will not work in the fittings.

I could really do too much investigating because it was late but it got me thinking.

What's usually the cause of this type a reading?

A Lost/loose/floating neutral?

 

[rapparee]

Yeah, it's obviously a broken neutral

 

[Lucien Nunes]

Yes, the neutral is disconnected / broken somewhere other than at the fitting. When you connect the meter L-N, the stray capacitance of the cable from the N core to earth allows a tiny current to flow through the meter, enough to make it give a reading but not one that indicates anything useful. The actual reading on the display depends as much on the input resistance of the meter as on the nature of the fault. An instrument with a very high input resistance would read the full 230V, a Martindale won't read anything at all. I hate the term 'ghost voltage' but this is an example of it, with the high resistance in the neutral rather than the line.

Regarding the term 'floating neutral', this is normally used to refer to a star-point such as the neutral bar in a 3-phase DB or distribution cable, that has become disconnected from the star point at the substation. In this case it would still be passing current but 'floating' about and changing in voltage as the loads vary. This is a problem because a heavy load on one phase can create a serious overvoltage up to 400V on another. In your case here, I would not call it 'floating' but merely a break or disconnect in the neutral conductor. (FWIW 'floating' also has another specific meaning of 'not referenced to earth' but that doesn't apply here.)

 

[gazdkw82]

Yes, the neutral is disconnected / broken somewhere other than at the fitting. When you connect the meter L-N, the stray capacitance of the cable from the N core to earth allows a tiny current to flow through the meter, enough to make it give a reading but not one that indicates anything useful. The actual reading on the display depends as much on the input resistance of the meter as on the nature of the fault. An instrument with a very high input resistance would read the full 230V, a Martindale won't read anything at all. I hate the term 'ghost voltage' but this is an example of it, with the high resistance in the neutral rather than the line.

Regarding the term 'floating neutral', this is normally used to refer to a star-point such as the neutral bar in a 3-phase DB or supply origin, that has become disconnected from the star point at the origin, and is still passing current but 'floating' about and changing in voltage as the loads vary. This is a problem because a heavy load on one phase can create a serious overvoltage up to 400V on another. In your case here, I would not call it 'floating' but merely a break or disconnect in the neutral conductor. A similar argument


(FWIW 'floating' also has another specific meaning of 'not referenced to earth' but that doesn't apply here.)

Brilliant information.

Thanks

 

[gazdkw82]

This is why this forum is invaluable. Quick educated responses.

Iv spoke with sparks who dont have a good thing to say about this forum. Maybe it's their persona or prejedgements....I dunno.

Personally I couldn't disagree with them more.

Because of this place I believe I have a greater understanding and knowledge than they do. They would rather struggle blindly than hold their hands up, ask for help and be thankful for it.

Thanks again

 

[Marvo]

With odd readings like this it helps if your tester is dual impedance ie it has a LoZ or low impedance range option usually around 3 or 4 kΩ. https://www.fluke.com/en/learn/blog/digital-multimeters/dual-impedance-digital-multimeters

 

[gazdkw82]

With odd readings like this it helps if your tester is dual impedance ie it has a LoZ or low impedance range option usually around 3 or 4 kΩ. https://www.fluke.com/en/learn/blog/digital-multimeters/dual-impedance-digital-multimeters

I'm not sure my meter does (metrel mi3000 easiplus) it does have plenty of options.

What's the theory behind this? Even if it had options I dont know which one to choose.

 

[Lucien Nunes]

A perfect measuring device doesn't change the thing it's measuring. In the case of a voltmeter, that means having an infinite input resistance so that it doesn't load the circuit being tested. For measuring voltage in most electrical work, it doesn't matter, any old voltmeter will do because the supply is the national grid which can spare a few milliwatts to operate the meter's input circuit, and in any case most digital meters have an input resistance in the order of 10 megohms which doesn't influence low power electronic circuits very much either.

But sometimes you actually want to influence the circuit. You want to put a bit of load on, to see whether it pulls the voltage down, to prove that there is a high resistance between the meter and the supply. That could be done by connecting a test lamp instead of the meter, as that has a much lower resistance, but it's handy to have the function built into the tester to lower its input resistance, so that you don't have to fiddle with other bits of kit.

 

[gazdkw82]

If you did attach a test lamp, you will still need to attach a meter to measure?

 

[Lucien Nunes]

Well, it depends what you are looking for. If you want to see that the 74 volts that the meter can detect is actually the result of trying to measure a 230V supply through the stray capacitive leakage of a cable, then you leave the meter connected and hook on your test lamp. The voltage indication drops from 74 to something much lower, maybe a few hundred millivolts, and you say to yourself 'OK, the resistance of the circuit is very much higher than the resistance of the test lamp, which means there must be a broken connection.'

Alternatively, you just hook on the test lamp and say to yourself, 'OK, it doesn't light up, there must be a broken connection.'

 

[GBDamo]

This is why this forum is invaluable. Quick educated responses.

Iv spoke with sparks who dont have a good thing to say about this forum. Maybe it's their persona or prejedgements....I dunno.

Personally I couldn't disagree with them more.

Because of this place I believe I have a greater understanding and knowledge than they do. They would rather struggle blindly than hold their hands up, ask for help and be thankful for it.

Thanks again

Pride and insecurity are very bad bedfellows.

No one knows it all but some feel they should and rather than ask will hazard a guess and hope.

This place seems to have calmed down a bit and the 'judgey' types tend not to be as vocal so feel free to ask stupid questions, I do.

 

[Mini-ohm]

Hi lads new on here! Id say Definately a lost neutral on the circuit somewhere!
a floating neutral is what i know as a broken neutral generally on a 3 phase supply which usually would blow up circuit boards and other sensitive electrical equipment as you would get funny readings and 400v live to earth etc! Something regarding odd voltages thoigh i have come across before and never really sussed out a few years ago was a lighting circuit working and testing as usual more than acceptable zs readings etc, but when the switch is turned off i was getting upto 48v can anyone explain this?

 

[Marvo]

I'm not sure my meter does (metrel mi3000 easiplus) it does have plenty of options.

What's the theory behind this? Even if it had options I dont know which one to choose.

This problem of taking voltage measurements and getting results that are numerically above what is considered a safe touch voltage but actually reflect voltages that are inconsequential in terms of the energy they can deliver has only been around for the last 3 or 4 decades since digital testers became the norm.

In the good ol days we all had analogue voltmeters which had an input impedance that far closer matched that of the human body. Because of this if you saw a reading of 60 or 70 volts you knew it had sufficient energy behind it to cause a significant current flow if you touched it ie in laymens terms you knew it was going to give you a whack.

Maybe rather than spending a fortune on a digital multimeter with a LoZ range go pick yourself up a decent analogue voltmeter. They're retro, all the cool kids are using them and they certainly get you noticed when you whip one out on site. I picked up a Triplett 310 when I was in Europe last year for a decent price and it's a pretty nicely built tester. Shop around and see what's available in your neck of the woods.

 

[Lucien Nunes]

Hello and welcome Miniohm.

when the switch is turned off i was getting upto 48v can anyone explain this?

You don't say where the 48V was measured, I assume across the lighting load SL-N. Obviously we have been discussing how a test lamp or Martindale tester presents a low enough resistance to drain off any leakage that enters the switch line and pull it down to a negligible voltage. But that relies on there being a low resistance circuit through the lamp when it is not lit, which is true of filament lamps but not always true of fluorescents and LEDs and I am going to guess that your lighting load was not filament-based.

Some types of load are non-linear (i.e. current and voltage are not proportional) in various ways and as an example, an unlit switch-start fluorescent batten without a PFC capacitor appears completely open-circuit up to the ionisation voltage of the glow-starter. With 48 volts across it, no measurable current would flow through it, therefore any leakage into the switch line from the permanent line (e.g. via cable capacitance) will register on your meter as the fluorescent will not sink it to neutral. At 120V however, the glow starter would strike and dump the leakage current.

Similar effects can occur with LED drivers, which are typically small switched-mode power supplies. They draw virtually no current until the voltage in their reservoir capacitor is high enough to make the circuit start up, so they won't completely sink any leakage. Some types flash periodically on remarkably low inward leakage currents as their reservoir capacitor charges up slowly to start-up voltage.

OTOH if you took the measurement with the lamps out, all bets are off!

 

[gazdkw82]

Hi lads new on here! Id say Definately a lost neutral on the circuit somewhere!
a floating neutral is what i know as a broken neutral generally on a 3 phase supply which usually would blow up circuit boards and other sensitive electrical equipment as you would get funny readings and 400v live to earth etc! Something regarding odd voltages thoigh i have come across before and never really sussed out a few years ago was a lighting circuit working and testing as usual more than acceptable zs readings etc, but when the switch is turned off i was getting upto 48v can anyone explain this?

I actually had something similar recently. I was pointed in the right direction by this place. It was basically induced voltage from the switch cables.

There is a pretty decent video from dses on youtube which shows how induced voltage works.

The way I dealt with it was to use a snubber which basically eats up the voltage.