if we confirm continuity between mains earth and the DUT intended probe earth point, are we safe to proceed testing using an earth referenced Oscilloscope? Is this a bullet proof way of knowing we wont create a hazardous earth reference with our probe earth?
I appreciate the need for differential probes when testing floating ground devices, but does this continuity test confirm we are safe to proceed without the need for differential probes?
I appreciate the need for differential probes when testing floating ground devices, but does this continuity test confirm we are safe to proceed without the need for differential probes?
Last edited:
Mostly yes, but only if you're doing this test while the DUT is not connected to the oscilloscope. In other words, the DUT's return must be earthed on its own and not through the oscilloscope's internal circuitry. Also keep in mind that there could still be a significant amount of inductance (and therefore HF voltage drop) between the utility earth and yours.
Some people use an isolation transformer along with an unearthed single-ended oscilloscope to make floating (differential) measurements. However, in such cases, only one channel of the oscilloscope would make any meaningful measurements, as the BNC outer body / ring would then be common to all.
I suggest that you draw a circuit diagram to understand the situation clearly before making any attempts at such measurements.
I suggest that you draw a circuit diagram to understand the situation clearly before making any attempts at such measurements.
Can't hurt to use differential probes anyway. Use them on their multiplier setting, too. Then use your oscilloscope's de-multiply setting. You actually get better results this way than without a differential probe - and end up not having to worry about blowing your scope.
thanks gents.. I hope this clears things for others new to scopes too. I was initially a bit puzzled, I blame some of the blabbering YouTuve video's for that!
I have read up on isolation transfo's, differential probes and lifting the scopes earth, the later I don't like for safety reasons. I plan taking the probe route.
I have read up on isolation transfo's, differential probes and lifting the scopes earth, the later I don't like for safety reasons. I plan taking the probe route.
It's the differential scope (not the differential probe) that is an alternative to the isolation transformer method. Regarding safety, the plastic insulation extends all the way to the end of the connector in decent probes, so that the user doesn't touch any metal while making measurements. Use only this kind if you plan on ground-lifting the oscilloscope.
However, note that if you want the differential probe, you will need a differential scope as well.
well thats not my understanding after reading up on probes and scopes over the last week while in Isolation
You can use any ground referenced scope with single ended probes to safely measure floating circuits simply by connecting both probe earths together then connecting each probe + on the test points, so we are not referencing earth at all, we scope math the difference. Sure this method may produce more noise compared to using differential probes or an iso transfo but at least we are not exposed to hazardous loops
Last edited:
I dont have a dedicated test bench so if I need isolation then probes are easier to change out than putting a isolation transformer on the AC input
Last edited:
That's not a differential probe but a measurement using two probes, not sure of CM issues of the method. The differential probe is one that is allowed to float (by a differential scope), giving you all channels as usual. See a video here:
https://www.tek.com/en/video/product-features/isolated-channels-safely-floating--accurately-testing
yes its not a differential probe, its single ended probes with floating earths. Its a simple method of eliminating ground or hazardous loops with single ended probes on ground referenced scopes. It allows you to reference earth and also probe across components safely..
Though I believe differential scopes are available, they are not necessary when using differential probes. The differential probe is designed to use on ground referenced scopes when you have potential for hazardous loops or you need to probe components without inducing a ground reference. The differential probe has an internal amplifier that creates the reference which is not related or common to the scopes earth. They are more accurate than the twin channel single ended method I outlined and they also help to minimise noise.
Isolation transformers on the other hand isolate the potential for ground becoming common to neutral when probing, this is achieved simply because you have no physical connection between the transformer primary and secondary windings.
Ground lifting the scopes earth input is no different to an isolation transformer, it breaks the potential for earth becoming common to the DUT though this method is dangerous and not recommend.
Last edited:
Feeding the DUT through the isolation transformer (and leaving the 'scope Earthed) used to be standard practice. Did I miss something?
Cheers
Cheers
Though I believe differential scopes are available, they are not necessary when using active differential probes. The active differential probe is designed to use on ground referenced scopes when you have potential for hazardous loops or you need to probe components without inducing a ground reference. The active differential probe has an internal amplifier that creates the reference which is not related or common to the scopes earth.
A regular (passive) differential probe does need a differential input scope to properly reject CM noise. Also, the two probe method's CM performance cannot be far from that of a single-ended measurement.
That's another way of doing the same thing.
.. yes I should have differentiated, thanks for adding active in, my ol' brain at it again 😵
Its well know single ended probe earth leads are a source of noise, so using 2 in the two probe method you'd have to expect more noise per channel, yes ?!?
Its well know single ended probe earth leads are a source of noise, so using 2 in the two probe method you'd have to expect more noise per channel, yes ?!?
Last edited:
in saying passive differential probe, are you referring to the regular passive single ended probe that comes with most scopes?
I searched but cant find a passive differential probe
Last edited:
Differential probes are ones that have similar terminals feeding a balanced cable. These are hard to find these days and most modern differential probes are active, with some capable of floating upto 600V / 1000V (CAT II / III) and 100x ratio.
There are photographs of both types in the following article. The differential probe does not have a core / shield structure, but instead gives the same status to both conductors. The impedance for both the conductors would therefore be more or less the same, and this improves HF performance. However, to obtain good CMRR, you also need a differential / instrumentation amplifier or a scope with a differential front-end.
https://www.emcesd.com/tt2007/tt010107.htm
With the two-probe method, the noise due to both the probes, both internal ADCs / amplifiers / acquisition all have to match each other exactly on a sample by sample basis, in order to get removed by the math subtraction. However, my best guess is that noise (being statistical) is mostly additive, and the resultant is thus going to be the vectorial sum of the individual noises.
There are photographs of both types in the following article. The differential probe does not have a core / shield structure, but instead gives the same status to both conductors. The impedance for both the conductors would therefore be more or less the same, and this improves HF performance. However, to obtain good CMRR, you also need a differential / instrumentation amplifier or a scope with a differential front-end.
https://www.emcesd.com/tt2007/tt010107.htm
With the two-probe method, the noise due to both the probes, both internal ADCs / amplifiers / acquisition all have to match each other exactly on a sample by sample basis, in order to get removed by the math subtraction. However, my best guess is that noise (being statistical) is mostly additive, and the resultant is thus going to be the vectorial sum of the individual noises.
Last edited:
There's a small difference that just came to my mind.
If one chooses the isolate the DUT, the isolation transformer VA rating needs to support the DUT's power consumption, which might be problem if the DUT is a high power appliance like an amplifier, SMPS etc. However, by choosing to isolate the oscilloscope the transformer needs to be rated for oscilloscope power (only), which is usually a small and predictable quantity.
However, if you already own an isolation transformer that is rated higher than any DUTs you're likely to have, then it maybe better (from the safety point of view) to isolate the DUT, while leaving the scope earthed.
Hope this is useful.
How about those hand-held, battery powered scopes? Are they not just like a DMM but with a graphical display? You dont worry so much about what the black lead is connected to when using a DMM.
For a dual channel measurement using such a beast (Fluke 123), I'm pretty sure the grounds of each probe still need to be connected to the same reference voltage - like both to the CT of your OPT primary, each probe tip to the plate or screen of the P-P tubes. The whole scope floats at 400V. Fluke even has a optically coupled computer interface (serial), where you could print the screen of your measurement. That's old stuff; unsure what modern HH scopes are capable of.
For a dual channel measurement using such a beast (Fluke 123), I'm pretty sure the grounds of each probe still need to be connected to the same reference voltage - like both to the CT of your OPT primary, each probe tip to the plate or screen of the P-P tubes. The whole scope floats at 400V. Fluke even has a optically coupled computer interface (serial), where you could print the screen of your measurement. That's old stuff; unsure what modern HH scopes are capable of.