Cutting earth connection to scope measuring mains voltages?

[treez]

Hello , I am testing mains powered SMPS's (+/-40V and 200W).
I need to measure voltages at the primary side of the switch mode transformer which provides isolation from the mains.

I am powering the smps as follows:-
Mains wall socket, 230VAC > Autotransformer > Power meter > SMPS

I cannot use an 1:1 mains isolation transformer in this path, as it affects the power meter readings -due to the isolation transformer's stray impedances. (just feel how hot an 1:1 mains isolation transformer gets even when its plugged in on no load)

Therefore i must simply cut the earth lead out of the scope's plug.

I cannot use a diff probe as they are £200+ and no company i've ever been to will buy them.



I have heard it said that due to the capacitance between the scope's "ground" and the mains hot and neutral, any measurements of waveforms taken with the earth lead cut out of the scope's plug will be inaccurate?
-is this true?


I saw this previous thread, though i could not find details on the accuracy with earth cut from scope's plug, or info on whether cutting the earth out of a scope's plug can damage the scope?
https://www.edaboard.com/threads/200985/

 

[FvM]

I know old engineers and careless engineers, but not many engineers who are both careless and old...

 

[treez]

Thanks,

Page 5, (bottom) of this
https://www.newark.com/pdfs/techarticles/tektronix/FFM.pdf

..says that cutting the earth out of a scope's plug makes it have a large capacitance to "ground".

How can this be so?....if you have just cut off the earth connection, then there is no "ground" in the scope, and so how can anything in the scope have a large capacitance to ground?

 

[FvM]

I see that the brochure elaborates the safety warnings about “floating a conventional grounded oscilloscope", so I don't need to further belabor the point.

Besides the general electrical safety issue, ground to mains dielectric strength and stray capacitance are particularly addressed.

As long as the summed ground to mains voltage doesn't exceed the regular voltage rating, dielectric strength isn't endangered. Connecting the ground to the DC- or DC+ terminal of a mains connected bridge rectifier should stay within the limits. But you shouldn't superimpose a high frequency switched voltage, e.g. connecting the ground to a H-bridge output.

The possibility to corrupt measurements by ground currents isn't specific to the floating ground situation however. It also applies to other situations where common mode voltages are present at the device under test, either because it has a (probably safe) floating voltage supply or the interferences are there although it's grounded.

 

[The Electrician]

Use an isolation transformer to power the scope. The one I have used has split bobbin construction and is rated for 5000 volt standoff between primary and secondary. The primary to secondary capacitance of this style transformer is much lower than the similar capacitance of the transformer in the scope.

It sounds like you understand the hazards of this; the ground and frame of your scope will be elevated to dangerous potential if you connect the ground clip of your probe to a high voltage node on the device you're testing.

 

[treez]

The above referenced document...
https://www.newark.com/pdfs/techarticles/tektronix/FFM.pdf

....discusses methods of doing floating voltage measurements.......but for some reason it fails to discuss the situation of reading the voltage across a high side sense resistor in a SELV (< 60VDC) situation, in an isolated power supply......surely if one has a grounded scope, and is measuring the voltage across such a resistor in an *isolated* power supply, then there is no problem?.....Though would you agree that a differential probe in this situation would give a less noisy scope trace of the waveform across the high side sense resistor?

 

[FvM]

but for some reason it fails to discuss the situation of reading the voltage across a high side sense resistor in a SELV (< 60VDC) situation, in an isolated power supply

Not surprizingly for a short application note, it also "fails" to discuss many other measurement problems which can easily fill a text book. Quite obviouly the AN focus is in promoting portable isolated Tek scopes.

Referring to the specific question, you have often additional connections to your circuit under test, some of them may create a ground loop. So I would be always careful when connecting the oscilloscope ground to a a highside node, even if it's only DC. If you have a differential probe at hand, it's a convenient method to measure the highside referenced voltage. In case of superimposed AC, it's probably the only way to get an undistorted measurement.

 

[Audioguru]

Post a video of you getting zapped.

 

[treez]

In case of superimposed AC, it's probably the only way to get an undistorted measurement.

there is a ripple current in the resistor (ie AC) ...so you believe diff probe measurement is the only way?, even if there is no other grounded equipment attached to the isolated circuit?

Basically , any resistor that does not have one pad connected to earth ground needs a diff probe to see its waveform?

 

[FvM]

With "superimposed AC" I was referring to large magnitude, possibly higher frequent voltage, e.g. at the center node of a half bridge. Some ripple shouldn'n be a problem.

You should particularly consider that you are loading the node where you attach the ground clip with the common mode noise of your isolated circuit which can be severe e.g. for s swiched mode power supply. If the node has low impedance, e.g. a +DC supply voltage, it's still no problem.

 

[THINKER_KAM_MAN]

just a suggestion.

is it not possible using a dual trace oscilloscope to use 2 similar probes
one on ch1 and the other on ch2.
set ch2 to invert. and set to add ch1 and ch2.
then connect the probes ground clip together
and probe using the two probe.

even if you use an isolation transformer the capacitance
of the transformer inter winding will give problems.
the cheapest way round the problem is to use a battery
operated oscilloscope if you got one.

 

[betwixt]

I agree with THINKER_KAM_MAN on this one. Provided the probes are rated to withstand the voltage you are measuring and you have a dual-beam or better scope, use "add & invert" to subtract the signals seen on two probes from each other. The difference is the voltage you are trying to measure. Leave the ground clips disconnected or linked to each other but nothing else. The high impedance of the probes should isolate their capacitance reasonably well. This method is safer because you can leave the scope grounded although I would never exclude the use of an isolating transformer on any test equipment just as an extra safety feature.

If your probes are not rated for high enough voltage, you can always add your own potential divider to each one, just be careful to keep the division ratio the same on both probes or take it into account in the Y gain settings for each channel.

Brian.