What's your definition of 'ghost voltage' ?

[Lucien Nunes]

I don't like the term 'ghost voltage' a.k.a. 'phantom voltage*' much, not least because it's rather loosely defined and probably means different things to different people. I also suspect that quite a few people use the term without really understanding what it is they are referring to.

My definition of what constitutes 'ghost voltage' is quite specific, but first, what is yours?

*OTOH 'phantom power' (as used to power condenser mics and other audio sources via the signal cable) is a different kettle of spooks and very clearly defined.

 

[SparkyChick]

I'm assuming you're talking about the sometimes spurious voltage readings we can get when testing for dead at various points in a circuit?

If so, I don't like the term ghost voltage... my personal choice is induced voltage because it is induced in the conductors by either induction, capacitance or a combination of both. It could also be residual voltage, from say a capacitor in a capacitive dropper or a filter.

 

[Lucien Nunes]

I'm assuming you're talking about

My point exactly. I know what I'm talking about, you know what you're talking about, but there's no guarantee we are talking about the same thing as there's no official definition.

It could also be residual voltage, from say a capacitor in a capacitive dropper or a filter.

I would not include this under the heading of 'ghost voltage'. It's the voltage due to residual charge in a capacitor, fair and square.

 

[snowhead]

I don't think I've ever use the term Ghost voltage but as S - Chick above, to me it's describing a measured (induced) voltage, but if you connected an appliance to it the voltage would drop to zero.

Not the same as a poor connection / failing switch or isolator where the circuit won't sustain a load, as that voltage comes from a true source.

There's no official definition because the term is too vague, almost Ghostly,,,

 

[DPG]

A voltage arising from incidental pickup by either inductive or capacitive effect rather than by direct and intentional connection.

That's mine, let me know what you think.

 

[SparkyChick]

I would not include this under the heading of 'ghost voltage'. It's the voltage due to residual charge in a capacitor, fair and square.

I agree, which is partly why I don't like/use the term ghost voltage because it's just too vague.

 

[ipf]

My point exactly. I know what I'm talking about, you know what you're talking about, but there's no guarantee we are talking about the same thing as there's no official definition.

Entirely agree. It might take a séance to fathom that one out as it's potentially spiritual.

 

[littlespark]

Ghost voltages. Measured in “caspers” apparently.

 

[Lucien Nunes]

A few people have mentioned capacitive or inductive coupling, what about resistive leakage into a conductor?

rather than by direct and intentional connection

This is very similar to one of my own criteria for qualifying something as ghost voltage.

 

[DPG]

A few people have mentioned capacitive or inductive coupling, what about resistive leakage into a conductor?


This is very similar to one of my own criteria for qualifying something as ghost voltage.

Resistive would be a different thing for me,as it would imply some sort of direct connection.

 

[static zap]

a)-- Not advisable ...
One that doesn't tingle
(turn you into a Ghost if touched !)
But detectable on delicate hang nail flesh.
------------------
b)... I would say vanishes with 22k ohm --
or am I not tripping here
c) Disappearing when drummond applied ?
(MFT test impedances are high enough to show phantom voltages)

 

[Lucien Nunes]

Aha, an actual number. Not sure what ratio 'vanishes' constitutes so can't convert that 22k into a source impedance, but a step towards my own definition.

To me the distinction between resistive leakage and capacitive leakage seems arbitrary. In any case if one wants to classify between deliberate and accidental sources, capacitance is often built-in but resistive leakage sources are just non-infinite insulation resistance, hence unintentional.

 

[DPG]

Good discussion Lucien.

 

[static zap]

So I need to hunt ghosts with an Oscilloscope , for their " Phase " relationship.
(My status will be spooked)
(or can we even squeeze this relationship out using phase rotation built into some 2 pole testers !)

 

[littlespark]

After my Casper joke.... I should have thought “who you gonna call?....”

But only if they’re registered with a ghost hunter competence scheme and listed on “trusted ghostbuster dot com”

 

[R-fur]

The National Grid Co have very specific terms for voltages on conductors and how they got there. I think they have four terms from memory, Trapped charge, capacitive coupled, induced voltage, cant remember the other one. They are talking about EHV circuits obviously the theory is similar.

 

[buzzlightyear]

funny you mention the post ,I had one to day on my megger no voltage present
and started do show OL ghost voltage , the last time it came up I rang megger the person on the other phone said how spooky , I started to singing goat busters bang me head against a wall.lol

 

[Darkwood]

It's obvious that none of you have ever seen a supernatural film, a ghost voltage is the thing that makes the lights flicker and the tv come on when it's not plugged in... seriously, where you been!

 

[dmxtothemax]

My definetion of ghost voltage is a measurable voltage that is of such a low current it can do no real work.

 

[telectrix]

it can do no real work.

a bit like an Aussie then.

 

[Lucien Nunes]

My definition. All three crtitera need to be satisfied to call it a ghost voltage:

A voltage present with respect to earth, or to some other identified point
a) on a conductor or component presumed to be or intended to be unconnected (but not necessarily safely isolated) i.e. floating.
b) derived from and having the characteristics of the supply (or one of the supplies) within the installation.
c) with a source impedance in the range 1MΩ - 100MΩ at freequencies of interest, or one decade outside this range under special circumstances

Point (a) excludes abnormal voltages on energised circuits due to a high resistance connections etc. and normal functional voltages that are not a recognisable supply voltage. In day-to-day electrical work many things are either connected to the supply or not connected, whereas in electronics most points have some voltage on them but are not directly connected to a supply. It does not make sense to me, to use the term 'ghost voltage' of a voltage expected to be present, just because it has a high source impedance, i.e. that it is not 'stiff'. This point also excludes low voltages of low source impedance such as differences between two connected earthing or bonding conductors due to circulating currents in a CNE.

Point (b) excludes kinds of voltages that exist on random conductors that have nothing to do with electrical wiring or are transient, e.g. triboelectricity (static), residual charge and dielectric absorption recharge, low thermal or electrochemical voltages etc. These usually have a signature that allows them to be immediately identified.

The lower resistance bound of point (c) puts the maximum current to earth on a 230/400V supply at 0.23mA, i.e. below the lowest threshold for permissible earth leakage on any appliance, while the upper bound ensures that a ghost of voltage 50V O/C gives a reading of at least 5V on a meter with 10MΩ input resistance. Working in electronics, especially where high-impedance circuits are involved, one gets a feel for source impedance without having to measure it, especially if one can reasonably estimate the O/C voltage and compare that with the actual reading.

Every isolated conductor on the planet has stray admittances (inverse of impedances). If it's large and well insulated, its susceptance even at 50Hz from its stray capacitance to earth will dominate. If it is part of a wiring system, the next most significant admittance is also likely due to capacitance, to other conductors. A floating conductor in a cable will often take up a potential between those of nearby cores and earth, and one can with a bit of logic and maths make all kinds of inferences from the readings that it gives: length to a break, position within a multicore, likely devices connected at the far end.

I've heard all sorts of nonsense about 'ghost voltage', such as 'it's not real voltage' and 'you can't measure it', whereas from an electrical point of view, it's just as legitimate a thing to measure and analyse as any other voltage.

 

[Ian Mills]

Here in France it is a big thing, because they use a lot of conduit with many cables in it from different circuits you get a lot of inductance in cables that should be 0 volts I have seen LED lights glowing all night from this

 

[MPTESE]

This may or may not be related or useful context for your question. In the film industry, there's a common term called "Ghost load". It's often used when dimming lighting fixtures with traditional dimmer packs. Sometimes the dimmer pack requires a minimum load to operate effectively and smoothly. If the load is minimal, often with LED lights, the solution is to attach a "ghost load". Sometimes we'll burn a 1k incandescent in line and point it away from set, and it solves our dimming problems.

Perhaps this is similar to what @dmxtothemax replied with.

 

[Lucien Nunes]

Same word, different meaning and context. Ghost load is a deliberate load that is hidden from view, ghost voltage is a potential from such high source impedance that it 'disappears' when tested with anything other than a high-resistance meter.

 

[HappyHippyDad]

Same word, different meaning and context. Ghost load is a deliberate load that is hidden from view, ghost voltage is a potential from such high source impedance that it 'disappears' when tested with anything other than a high-resistance meter.

I remember a while back I mistakenly said ghost voltage is 'not real' and you instructed me otherwise. I still can't quite get my head around why we can't get a shock from it if it is 'real'?

I tested a 19 core cable and got around 200V between one of the conductors (which was completely disconnected) and earth when tested with my MFT. I then tested with the drummond and got no light at all (so <50V). Is this useable voltage and if so why don't I get a shock? Is it 200V as the MFT says or is it <50V as the drummond says?

 

[DPG]

Technically speaking you could get a shock from a 'ghost' voltage. In fact consider the ghost voltages and currents available on transmission lines.

 

[Lucien Nunes]

Voltage is potential, the ability to do work by driving a current through a resistance. But most of the effects that electricity can produce, like heating an element or giving you a shock, depend on the value of the current. If the current is drastically limited in some way, then even if the source voltage is high, the effect is also drastically limited. In the case of ghost voltage, the current is limited by a very high impedance between the source (a truly live conductor) and the conductor οn which the ghost voltage is detected.

If that current-limiting factor (e.g. a stray capacitance from a truly live core of a cable to another that you are testing) has an impedance of one megohm, on a 230V circuit no matter what happens the current available to the touch is limited to 230/1,000,000 = 0.23mA which will not give you a shock, or light a lamp, or give an indication on a traditional Drummond tester. However, 0.23mA is plenty to operate a digital multimeter's voltage range, and it will give you a reading in the order of 200V. It's really 200V while it's there, but connect a lamp (which has a much lower resistance than the meter) or bridge it to earth with your hand, and the voltage collapses because of that very limited available current.

When I have more time later I'll set up some examples and examine them as voltage dividers.

 

[HappyHippyDad]

Voltage is potential, the ability to do work by driving a current through a resistance. But most of the effects that electricity can produce, like heating an element or giving you a shock, depend on the value of the current. If the current is drastically limited in some way, then even if the source voltage is high, the effect is also drastically limited. In the case of ghost voltage, the current is limited by a very high impedance between the source (a truly live conductor) and the conductor οn which the ghost voltage is detected.

If that current-limiting factor (e.g. a stray capacitance from a truly live core of a cable to another that you are testing) has an impedance of one megohm, on a 230V circuit no matter what happens the current available to the touch is limited to 230/1,000,000 = 0.23mA which will not give you a shock, or light a lamp, or give an indication on a traditional Drummond tester. However, 0.23mA is plenty to operate a digital multimeter's voltage range, and it will give you a reading in the order of 200V. It's really 200V while it's there, but connect a lamp (which has a much lower resistance than the meter) or bridge it to earth with your hand, and the voltage collapses because of that very limited available current.

When I have more time later I'll set up some examples and examine them as voltage dividers.

That's a great explanation Lucien, thankyou, I understand it alot more now. I guess it's the same with the 1000V IR test on the MFT. I remember the tutor giving me a shock with it, but it's not really that bad which I assume is due to the 'current limiting factor'.

 

[Vortigern]

Speaking from an installer/tester (be I ever so 'umble sir) the more technical aspects you bring up do not usually press to the front of considerations in such work. In common parlance ghost voltage is a voltage that is not derived directly from the DB. It is generally understood to be an induced voltage from other cables close to the one being tested and running along with it. It is important on a pragmatic level to distinguish between the voltage which is inducted and the voltage arising from a fault in the wiring/installation. So we introduce a higher load in the tester to see if the voltage is sustained under load which give a pretty sure indication of which of the two we are dealing with, fault or inducted voltage. It is of note that in the instruction manual for the MTL20 drummond tester that this voltage is referred to as interference voltage and phantom voltage and perhaps where the term ghost voltage came from.
The finer point you have brought to scrutiny while giving food for thought as well as education would rarely feature in on the ground (so to speak) day to day work. Not denigrating what you contribute but finding practical applications for it would be a challenge. Notwithstanding the foregoing I would encourage your suggestion of some worked voltage divider simulations of such situations as a picture says a thousand words and do sincerely look forward to the same!

 

[Lucien Nunes]

However, faultfinding and analysis are an art/craft/science that can be taken to whatever level one likes. Even when confronted with simple, everyday problems, you can still benefit from the use of advanced techniques and insights because they can deliver devastating speed and efficiency.

Example: Some here would say that voltsticks are the work of the devil and have no place in a proper electrical toolkit. But, on one occasion, using just a voltstick, I was able to locate and identify which strapper cable was not making a good connection, in which switch, in a 2-way + intermediate setup in about 30 seconds without taking any faceplates off and hence without even visiting the CU. I was then able to isolate the lighting circuit, remove the requisite switch, reterminate the loose conductor and put it back together with a total downtime of maybe 2 minutes. Using only the so-called dangerous voltstick, aided by my knowledge of its method of operation and parameters, and a knowledge of the shapes of the metal components inside an intermediate switch and the shapes of the electric fields they will produce in the plane of the front plate. Simple problem + advanced method = happy customer + £300/hour. QED.

 

[Dave OCD]

How much ??!!

 

[Lucien Nunes]

A notional £300/hour, based on being able to complete a faultfinding and repair job from start to finish in 6 minutes, with a minimum charge of £30 per job.

My point being that although I hear what Vortigern says and I get where he's coming from, many jobbing sparks might still easily recoup the time invested in mastering some troubleshooting skills.

 

[Vortigern]

Well I did earn that much the other day (notionally!) when asked to trouble shoot lights not coming on and emergency lights non operational in a commercial set of toilets. They had looked at the problem and no one could solve it. Twenty secs looking and I solved the problem. That is because the whole system is in my head like a computer simulation and I reasoned it out pretty fast it was actually quite simple. Some scamps had got up to mischief and turned out the stubs left on the pull switch (left after conversion to sensors) on a ceiling four meters high, don't know how they managed to get up there. So there are more ways...etc. I do accept your posit that a higher level technical knowledge can increase the chances of discovering what the problem is but most of my problem solving is straight forward shorts, disconnections failed components pretty straightforward. But I love to hear your thoughts on some of the more tech aspects, not that I might add I fully understand it all.
P.S. but that job you did....was in your own house though?

 

[Lucien Nunes]

They had looked at the problem and no one could solve it.

This.

IMO, every spark should be able to resolve, with certainty and efficiency, simple problems with things not connected up right, disconnects, shorts, the things you mention. But it took someone with a bit more imagination and skill to do so.

No, that broken strapper was not in my house, I don't have any intermediates, my wiring isn't that posh. OTOH I don't touch domestic wiring commercially; I just fixed it for a customer while I was there for something else.

I should do some faultfinding videos. There are many separate aspects to good faultfinding, I could write a very thick book about it but who would read it? So much more fun to find out for real why the jukebox plays the wrong record or the washing machine sometimes stops at the second spin, or the spatial interpolator aperture function looks a bit off.