Measuring Mains Voltage with Oscilloscopes

Problems with measuring mains voltages with an oscilloscope, the correct and safe way to do it, and other dangerous methods that must never be used.

► Support this channel:

► Social Media:

► Contact info, sending stuff in: etc.:

► Website - More on this video and many other subjects

[ianharrison]

Thanks JW. Perhaps the clearest explanation of correct ‘scope/mains practice I’ve heard.

[daveb]

*My six year old daughter loves this channel. She especially likes learning about electrical code and regulatory bodies. You should make a second channel called JWKids which would focus more on the laws governing electrical code*

[m.]

Clear and precise explanation, going into detail that is easy to understand.
Thanks for your effort to share your knowledge with the rest of us John.

[hoggif]

Back when I made an electronic work bench wiring it was required that any open device you test/adjust must not be referenced to earth. And touching any single point in open device should be made safe.
For measuring I used an (mains isolated) transformer to power AC load. However, when you connect a scope with earth referenced led connected to earthed case, device under test becomes mains referenced. So, scope and other measurement devices went behind a separate isolation transformer. At that point you can safely touch any single point of a powered circuit with scope connected (as required).
That setup worked very well but it was bulky and heavy. I had 230V/16A rating for device and a variac added for testing at different voltages. That pack needed wheels under it to be able to move it by one person. Differential probes were out of hobbyist reach at that time for any reasonable bandwidth.
I think back then it was recommended to make tables etc from non-conducting material. I think isolation requirement for professional setups was 10kVac, much more than 4kV isolation protecting the normal product user. I actually tested my isolation for 10 or 12kVac.

[BjornV]

A few sidenotes.
6:25 Those probes or "external box" you mentioned, are called differential probes, but differential probes are NOT isolated (There are fully isolated probes, but those are even more expensive then a differential probe). With a differential probe there is still a high resistance between both input + and - leads, and between each lead and earth, but no isolation like a real isolation transformer.
8:25 By disconnecting the earth to the scope has also 2 major side effects on scopes wich contain a EMI filter on the mains input inside the chassis. If you don't connect a earth to this EMI filter, then 50% from your main voltage (with a very low current) will leak through the safety caps from the phase (hot wire) to the chassis (and negative part of your BNC connectors). Even when there isn't any probe connected to the scope. In the UK with 110a120V AC, there will be 55a60V AC on the chassis and BNC connectors. In countries with 240V AC, it's 120V AC what is present on the chassis and negative part of the BNC connectors.
When earth is connected to this EMI filter, this low current leakage is not a problem, and will be leaked to 0V.
A second side effect by disconnecting the earth to a scope with EMI filter inside, is that the produced measurements on the screen are faulthy, because there is no 0V reference point, this 0V point is now 50% of the mains voltage.

[Markus]

Very clear and thorough explanation. I 'sort of' understood this before, but now I 'really get it.' Thanks.

[sumilidero]

I always used scope probes with removed ground clips (it just unhooks off the probe) thus leaving scope ground as mains earth, and put ground-clip-less probes to 3 phases . Never had problems with too high voltage or transients (scope and probe impedances are high enough) altough connected mostly on the input of AC/DC and AC/AC 4Q converters. Of course scopes were digital not the bulb ones. For single phase small power devices, always isolation transformers. Also, never tripped RCD refering probes to earth as their impedance is high enough and the current is very little. Unlucky that all newbie technicians have no idea about the problem you perfectly explained, they just wanna stow the probes to the outlet like normal multimeter and watch a fun show on the screen. In schools they always probe some low voltage isolated things and reading only 400V input they think is enough to not to worry about anything. Great vid!

[madbstard]

My second lesson in two days on using a scope on mains - Play With Junk also talked about the dangers when he was probing a linear power supply. But your video gave far more detail on the different scenarios.

[kenfoland]

If you are new to isolation transformers, research the subject thoroughly. Before you buy one, understand that not all isolation transformers are intended to be used in a test bench application. Some will pass the mains earth to the secondary coil by bonding the neutral lead to the primary earthing. This type is not suitable, and will require modification before use. Also, a cheap variac autotransformer alone will not work. It will not provide Galvanic isolation. Therefore, it still will require the use of a properly mains earth isolated isolation transformer.

[UpcycleElectronics]

My favorite 'other' option I've seen used is to simply add a loop or two of loose insulated wire to any magnetics already present in the circuit. While the turns ratios and actual voltage levels of the circuit will still be unknown as far as the scope readings, this does create a small scaled, isolated voltage that will allow you to see the AC waveform.

[peckelhaze]

There are other videos on YouTube regarding this topic but this is the best I have seen. EEVblog is good too but not quite as clear in my opinion. I have seen people suggest disconnecting the earth which as JW has said is ridiculously dangerous.

[julianhicks]

Hi John, Just watched this video. You can use the two channels of the scope and then there is a setting to display the difference between the two channels. Then you don't go anywhere near the external connection. No need to connect to the outside connection of the BNC connector. Just plug one channel to one side of the supply and the other channel to the other side.

[Robonthemoor]

Find this man ace, wish I had thought of been an Electrition, , love your work immensely 😀

[jonpotato]

Very accurately explained! There are so many ways you can destroy an osciloscope or the equipment you are probing. This video should be watched by anyone thinking about buying and using an oscilloscope. It should be watched BEFORE using it. Unfortunately for my cheap China USB scope, I watched your video afterwards.

[ginbot]

I built a circuit that used high value resistors for a voltage divider, then an isolation amplifier to keep the high voltage away from the oscilloscope. I find it especially useful for measuring non-sinusoidal waveforms like modified sine wave inverters.

[tarunarya]

Love your videos with clear reasoning and good solutions.

[ManchesterBlackSheep]

Fabulous video; beautifully, clearly explained. thank you very much. The "fairly fatal" amused me! :D

[metalsnapper]

Hi John. Love watching your videos and you indeed one very clever bloke! 👍🏻. I’m looking for a handheld oscilloscope to be able to see the waveforms for 230v AC and also DC voltages up to 48V. This would be for testing the outputs on portable generators & DC to AC inverters. These will be installed on boats as some of these generators and inverters don’t produce and very nice sine-wave and cause issues for the boat owner. It would be really nice to be able connect a scope up to these and see if they are actually producing a nice waveform so I can 100% prove that it is their inverter or generator. Some of these inverters do actually say they are modified sine wave but some don’t. I’d like to know which handheld meter can do this safely and would I still require a voltage divider. Many thanks. 👍🏻

[billinski]

As always, thorough simple arguments and solutions.

[NivagSwerdna]

Very good. I did not know about this at all! And now... I need to go shopping for an isolation transformer!