Question about grounds in circuits

[CraigAB]

Hi,

I have been watching Dave's "How not to blow up your Oscilloscope" and am wondering:

Q1: As long as the grounds are all connected everything is ok? (A common ground plane / reference ground?)

Q2: If the ground lead on the scope is connected to a circuit (with an isolated main) that is ok? (The scope is using the Cct ground as the reference?)

Q3: If the probe ground is not connected to any ground, then it's reference is the mains ground reference?

Regards,

Craig

[tggzzz]

Partially. There is no such thing as a "ground" where everything is at the same potential. There are only approximations useful in some circumstances.

One example: if caught in an open field in a thunderstorm, crouch with only your feet on the soil and your feet together. That reduces the chance a nearby strike doesn't cause currents that find it easier to go through your legs than the soil.

Another example: if you manage to make a purely resistive wire, you will have made your fortune. All wires currently on sale come with a free inductor (rule of thumb is 1nH/mm).

And, of course, all conductors are also antennas.

For how to use scopes effectively, have a look at the Tektronix references in https://entertaininghacks.wordpress.com/library-2/scope-probe-reference-material/

And.welcome to the forum

[Jwillis]

If oscilloscope is Earth Grounded, ground clips must be at the same potential when using multiple channels.

It does not matter if oscilloscope or DUT(Device Under Test) or both are Isolated from Earth ground, ground clips still need to be at the same potential. This becomes very important when measuring a DUT with very high voltages because the ground clips are Not isolated from each other for each channel and ground potentials can very from one location to another.

If  oscilloscope has no earth ground and/or DUT has no earth ground and no ground clip is attached to DUT there is no reference. A proper measurement cannot be taken.
The earth ground is used for safety and should not be used as the reference. Earth ground and the probe grounds are separated by high resistance and would give an inaccurate measurement.

[tggzzz]

It does not matter if oscilloscope or DUT(Device Under Test) or both are Isolated from Earth ground, ground clips still need to be at the same potential.

As i'm sure you are aware, but a beginner might not be, it very much matters from the point of view of safety. Never disconnect a scope's protective mains earth, since "floating a scope" has killed people, killed scopes, and killed UUTs. FFI, see the safety and praxis sections of https://entertaininghacks.wordpress.com/library-2/scope-probe-reference-material/

Be aware that some posters will make statements equivalent to "I've run into a road without looking, and I didn't come to any harm. (Therefore you can too)". Some risks are worth taking; others aren't.

[CraigAB]

So I think I understand it:

If the scope and the Device under test have the same reference ground, you don't need to use the scope probes earth lead.

If the scope is using the House ground and the device under test has an isolated ground (Powered by a battery), then you need to use the lead on the probe so that the scope is using the same reference ground as the device under test.

As I (hopefully) understand it, the biggest problem is connecting the scopes earth lead, on the device under test, to something that is thought to be ground, but is in fact some (high) voltage? And Zap/bang/smoke.

Or touching your hand to something inside the device under test to (high) voltage and creating another (shortest) path to ground (earth).

Thanks for your patience.

[ArdWar]

If the scope and the Device under test have the same reference ground, you don't need to use the scope probes earth lead..

Technically yes but doing it that way will create a huge return loop and you'll pick every interference, ground loop, stray voltages, parasitic inductance, capacitance and whatnot along the way. On typical house wiring a small signal will be drowned by line frequency hum. High frequency or fast signal probably will be completely lost.

If the scope is using the House ground and the device under test has an isolated ground (Powered by a battery), then you need to use the lead on the probe so that the scope is using the same reference ground as the device under test.

Yes, and you should do that anyway for like 99.9% of cases. You need your measurement loop as small as possible to properly capture fast signals. Not using the ground lead is almost never actually done, and I can't even think of an use case for that beyond curiosity. If you want to probe a non ground referenced non floating signal you should use isolated differential probe. Use the proper tool for your own safety and measurement integrity.

As I (hopefully) understand it, the biggest problem is connecting the scopes earth lead, on the device under test, to something that is thought to be ground, but is in fact some (high) voltage? And Zap/bang/smoke.

Yes, if the high voltage is not floating with regard to scope's ground.

[CraigAB]

Technically yes but doing it that way will create a huge return loop and you'll pick every interference, ground loop, stray voltages, parasitic inductance, capacitance and whatnot along the way. On typical house wiring a small signal will be drowned by line frequency hum. High frequency or fast signal probably will be completely lost.

So to reduce that interference, I could use a ground lead from the scope to the Device under test's ground?

[EPAIII]

So I think I understand it:

If the scope and the Device under test have the same reference ground, you don't need to use the scope probes earth lead.

...<snip>...

That is only true for DC and low frequencies. If you are probing high frequency signals, then the length of the ground connection can make a big difference. This is why the ground wire is brought out at the probe end of the probe's cable. And why a short wire is used from there, and often the simple spring wire connection which is pressed into a ground connection on the PCB within millimeters of the point where the signal is being probed.

Generally speaking, the upper end of audio frequencies is about the limit for not using the probe's ground connection.

AND

...<snip>...

As I (hopefully) understand it, the biggest problem is connecting the scopes earth lead, on the device under test, to something that is thought to be ground, but is in fact some (high) voltage? And Zap/bang/smoke.

...<snip>...

That is exactly correct. You do not want to accidentally ground a point in the circuit that is not at ground potential or that can not be forced to it (as in a device powered by a battery or a floating power supply). Professional, well designed PCBs will have numerous ground points clearly marked at various, needed points around the board. Generally speaking, it is safe to connect a scope's ground clip to any one of these.

If the PCB is not designed and built to that level of professionalism, then you need to be extra careful. This would be a good place to use a good quality VOM, starting with the highest Voltage setting if it is not auto-ranging.

AND

As for touching things with your hands, TURN THE POWER OFF FIRST and WAIT A MINUTE OR TWO for the caps to discharge.