Tech Talk - What Is A Ghost Voltage And How To Test It

[@owaisakhter2673]

You explained far batter than anyone on KZbin

[@EriksElectronicsWorkbench]

Thanks for the nice compliment!

[@mathiasanders3946]

I am working for now nearly 30 years as an electrician and this is a common problem aspecially in harsh or wet envoirements and here where we use 230/400V. For example machines and motors going on and off, not by the ghost-voltage itself but the relais are switched on and off by this phenomena. And this is the reason why i carry always a two-pole-tester with switchable load like the FLUKE T110 or a BENNING DUSPOL etc. greetings

[@robharley9838]

Very interesting subject Eric, thanks for exploring it! I've never encountered the concept. None of my DMMs have a low Z setting, but it is fun food for thought as to how one might measure the ghost other ways. Appreciate your work! - jrh

[@timharig]

In order to remove the "ghost" voltage, there must be a load impedance significantly lower than the output impedance of the "ghost" voltage. That lower impedance can but does not have to be provided by the meter itself. You can simply add an appropriately sized resistor in parallel to the meter to provide enough load to pull down the "ghost" voltage.

[@robharley9838]

@@timharig Thanks for the insight timharig! Impedance is still a little bit mysterious for me in some situations, but discussions like this help with my understanding. Appreciate it.-jrh

[@timharig]

@@robharley9838 Don't let the word scare you. For the purposes of this video, since there is only one frequency involved and we are not dealing with reactive components, you can treat impedance just like resistance. Ohms law for impedance is the same as for resistance: V= IZ Serial impedances are additive: Zt = Z1 + Z2 + Z3 + ... Parallel impedances follow the same rule as parallel resistances: Zt = 1/(1/Z1 + 1/Z2 + 1/Z3 + ...) The voltage dropped in series impedances drop in proportion to their impedances: V1 = V*Z1/(Z1+Z2) V2 = V*Z2/(Z1+Z2) etc. If you have a 120V ghost voltage with an output impedance of 2Meg ohm and you connect a meter with a 10Meg ohm input impedance, then the voltage that the meter measures will be the drop across its 10Meg ohm input: V = 120*(10Meg/(10Meg + 2Meg)) = 100V If however, measure the same 120V@2Meg Ohm ghost voltage with a meter that only has a 100k Ohm input impedance, then most of the voltage is dropped across the ghost voltage's impedance and the meter will only measure: V = 120(100k/(100k + 2Meg)) = 5.7V If you have a 10Meg ohm meter and you want to reduce its apparent input impedance, you could take the 100k ohm resistor and connect it in parallel with the meter. Then the meter's apparent impedance is: Zm + Zr = 1/(1/100k + 1/10Meg) = 99k ohm If you wanted to further remove the ghost voltage, then you could use an even lower value resistor. Assuming a 1k ohm resistor: Zm + Zr = 1/(1/1k + 1/10Meg) = 1k V = 120(1k/(1k + 2Meg)) = 60mV Note however, that if the voltage being measured is NOT a ghost voltage and the output impedance of the REAL voltage was say 0.1 ohm, then the current flowing through such a small resistor could be significant: V = 120(1k/(1k+0.1)) = 120V I = 120/1k = 120mA That is enough to make a small resistor warm and enough to interfere with the operation of lower power circuits -- which is why volt meters usually try to have as high of an input impedance as possible. Also note that if we ARE dealing with reactive components (inductors [coils] and capacitors) then the equations above still work; but, we have to treat impedance as a complex vector and we need to be concerned with how the impedance of reactive components changes with frequency.

[@robharley9838]

@@timharig Many thanks for the lesson! OHMs law rules. I'll be pondering this for sure. -jrh

[@johnfitzpatrick2469]

Greatly appreciate your lesson. Australia supplies 240v L N G * I will seek a meter with Z * Will check for ghost V 💫🇦🇺

[@waynegram8907]

An examples of various circuit boards having Ghost Voltage Sources? What can be a Ghost Voltage Source? Most DVM meters don't have a Low impedance mode, can't you use a 1K ohm/5 watt or 10 watt resistor across the DVM meter port leads to put a LOAD on the high impedance DVM meter?

[@EriksElectronicsWorkbench]

My experience is ghost voltages are more likely on long runs of parallel wire, induced and picked up by nearby power transformers, and by capacitive coupling effects of nearby conductors. A typical circuit board on its own probably won't have ghost voltages but of course anything possible. Yes you can load a high impedance DMM with a parallel resistor of appropriate size and eliminate or greatly reduce the ghost voltage.

[@waynegram8907]

@@EriksElectronicsWorkbench The formula is ohms law to compute the coupling resistance is Ghost Voltage / by current = coupling resistance. The Ghost Voltage is saying that there is coupling capacitances/coupling resistance? The Ghost Voltage is leakage current & Ghost voltage causes higher signal to noise ratio?

[@EriksElectronicsWorkbench]

@@waynegram8907 it is not as simple as ohms law because you don't know the actual ghost voltage, because the simple act of measuring the ghost voltage alters it to a lower value. Even a high impedance DMM with 10 or 11 megohm impedance is loading and lowering the ghost voltage by some amount. If you want to know the resistance you'd need to use an insulation tester that applies a known high voltage and measures the current thus giving the resistance. Looking at a ghost voltage source, readings in the many gigaohms or even teraohms would be expected.

[@waynegram8907]

@@EriksElectronicsWorkbench Yes the Insulation Tester or Megger will measure the wire under test with referenced to ground or the wire under test with referenced to another wire in the harness. The general rule of thumb is the LOWER insulation resistance = HIGHER Ghost Voltage?

[@EriksElectronicsWorkbench]

@@waynegram8907 yes as the resistance goes lower the apparent ghost voltage will be measured higher. If the resistance falls low enough you will have a serious shock hazard.

[@izzzzzz6]

something i don't understand with isolation transformers. The voltage is still the same so must still be dangerous but you are also isolating from the RCD protection circuit so how does that male it safer? I guess if you touch both live and neutral on the other side of an isolation transformer you are more likely to just burn yourself rather than have the current flow through you (possibly essential organs) to ground but if you touched one of eack terminal in one of each hands then you would be at a greater risk.

[@EriksElectronicsWorkbench]

Yes the secondary of the isolation transformer is just as dangerous as the normal AC line voltage. If you contact between both ends of the transformer's secondary you will be shocked. The protection or isolation the transformer offers is that there is no longer a ground reference which reduces the shock hazard. In other words, if you contact ONE SIDE of the transformer's secondary (and assuming the transformer and wiring is set up correctly) there is no path the current can take through your body to ground.

[@Happymacer_P]

Yes, but the second important benefit of an isolation transformer is that on mains powered test gear, the probe return wire is usually connected to ground. By making your test gear float you can protect both your test gear and yourself by preventing a live-earth short during testing.

[@sharg0]

Now to check if any of my (cheap) multimeters have a low impedance setting, I doubt it.

[@migsvensurfing6310]

Just add a resistor across the terminals of the meter.

[@almightytreegod]

@migsvensurfing6310I was wondering if this might work, since I don’t have one either. I’m guessing because it’s in parallel and not in series, that’s why it lowers the impedance instead of increasing it, right? What value/rating resistor should be used? Does it need to be a high wattage or with it being parallel that doesn’t matter?

[@Janktzoni]

Thanks for this refresh, very clear and well explained! The use of metal boxes and conduits in domestic situations is making me somewhat nervous. The reason is that so many things can go wrong with these and the 'user' cannot safely or visually detect it is wrong or not without measurement equipment and the knowledge on how to use it. I believe the lower metal box itself on the isolation transformer should be earthed too. After all it is a metal box and the internal wiring can come loose and put the box on a dangerous potential. _Only the box_ and NOT the socket, as you correctly point out. (KZbin police mode on) Constructive criticism: Although you correctly explain that on the analog meter the red scale is applicable for AC, you're actually pointing to the black DC scale multiple times. The AC scale is not completely linear, the first 10% is shorter than on the DC scale. Admitted, this doesn't make a significant difference in this context and I would probably have made the same 'mistake'. (KZbin police mode off ;-)

[@EriksElectronicsWorkbench]

The metal box with the isolation transformer outlets should not be earth grounded. The reason is, suppose a wire comes loose and contacts inside the metal box. And suppose the box were grounded. Now one side of the isolation transformer has a ground reference and a shock hazard exists from the opposite secondary winding end to ground which completely defeats what the isolation transformer was supposed to protect against. With the box not grounded, suppose a wire comes loose and contacts inside the box. A shock would only occur if you touch the metal outlet box and at the same time the opposite secondary winding end which is very unlikely to occur (but of course not impossible). Regular testing of the whole isolation system for such faults is needed and if you prefer a plastic outlet box can be used. I will try to be more clear when pointing to the meter scales as to which tic marks I'm using but I did mention the AC voltage is taken on the red scale.

[@timharig]

The entire POINT of an isolation transformer is to isolate the connection from ground for safety. That way you cannot get shocked by grounding yourself out accidentally. The only way you get shocked is by bridging the two legs of the transformer. If one of the wires from the isolation transformer contacts the metal box, it's fine because the only path back to the transformer is through the other leg of the transformer. However, if you connect ground to metal box and one of transformer legs also touches the metal box, then ground becomes a path back to the transformer and you can once again electrocute yourself by accidently grounding yourself.

[@MartinCail]

I don't understand what you're saying at the 5:20 mark. I also don't understand how this example has a real world application. Ultimately what I'd like to know is how to test for ghost voltage in an old metal box with now ground wire.

[@EriksElectronicsWorkbench]

At the time mark you mention I am showing how you could be shocked if you were to come in contact with the hot AC line. The voltage potential exists between the hot AC line and earth ground. If you measure a voltage with a DMM that has the low impedance mode as I show, and the voltage drops away to near zero, but the voltage remains with a DMM having regular high impedance then it was a ghost voltage.

[@MartinCail]

@@EriksElectronicsWorkbench how do I test for ghost voltage in an old metal box with no ground wire? I work on old houses that just have a white and a black. I just bought a dmm with the low impedance setting. can I use the metal box itself as the "ground?"

[@EriksElectronicsWorkbench]

@@MartinCail the metal box may or may not be at ground potential depending on wiring. The ghost voltage would be between the metal box and the AC line(s) if it is present at all. From the black to the white wire is the AC line voltage, not ghost voltage so be careful. Take the volt reading in the normal AC measurement mode then take the reading in the low impedance mode. If the voltage disappears in the low impedance mode then it was a ghost voltage. If the voltage remains in the low impedance mode it is a live AC line or live AC between the points you are measuring meaning a shock hazard exists.

[@davidfalconer8913]

Some older small LED lamps exhibit a ( ghost ) glow when turned off ... this is due to the leakage capacitance of the switch drop wires ... many folk find this annoying ... but ... these amazing lamps give an ( almost free ? ) NIGHT LIGHT ! ! ... ( handy to avoid tripping over the cat at night ? ? ) ..... ( tried - n - tested ) .......... DAVE™🛑

[@EriksElectronicsWorkbench]

Yes LED's require very little current to operate (compared to an incandescent bulb) and with necessary current limiting resistors they are fairly high impedance loads.

[@Janktzoni]

Free: unfortunately not, while it is very little energy it is measured by the utilities meter.

[@davidfalconer8913]

@@Janktzoni I did say : ALMOST free .... this tiny glow would probably cost about ( ? UK , £1 ) for the whole year , which is VERY affordable ( Ha - Ha ! ) ........ DAVE™🛑

[@paulstaf]

My Fluke 87V has a LoZ setting, however it read 120 volts through my isolation transformer to ground. When I used my Fluke 117 that also has a LoZ setting, the reading was close to zero, congruent to your video. I connected a light bulb through my isolation transformer to ground and it did not light. Any idea why the Fluke 87V did not read correctly?

[@paulstaf]

Nevermind, my 87V does NOT have LoZ setting, it has an Lo setting for "Low Pass" filter. My 117 has an LoZ setting.

[@EriksElectronicsWorkbench]

The Fluke 87V doesn't have a Lo impedance setting. The LO feature in yellow marking by the AC volts range is actually a low pass filter useful when measuring noisy devices like variable frequency drives for motors. The standard way to show the two functions is that the low pass filter is shown by the marking LO with the flat and sloping line over the letters, the Lo impedance feature is shown by the marking LoZ.