OScope Safety Question

[KungFuJosh]

Hi, I'm not quite a newb, but I am when it comes to oscilloscopes. I've read a lot of mixed opinions regarding safety, and I would like to see if my setup makes sense, and whether or not I should bother with a differential probe.

I have a DS1054Z connected on a surge protector to a GFCI outlet. I have an isolated transformer (Toroid ISB-060W) with a verified floating neutral, also connected on the same surge protector to the same GFCI outlet.

Any DUT goes to the IT, and I don't use any adapters to defeat the earth. I'm generally only testing audio equipment like tube guitar amplifiers, and microphones. I may test guitar pedals too, but those are low voltage and I can battery supply them. I also have a Tekpower TP3005P bench supply for DC but I haven't tested that yet to see if it's isolated or not.

Generally speaking, I probably designed and built 99% of the amps or mics or pedals I'm testing with my scope. I'm obviously very family with the circuits and I'm not going to connect the ground clip to B+ and blow anything up.

All that said, what do you think? Is my setup sufficient / does it make sense? Should I bother getting a differential probe, and if so, what models should I look at?

Thanks,
Josh

PS. Yes, I've seen the "How NOT To Blow Up Your Oscilloscope!" video.

[David Hess]

That is how I do it. The idea is that with the isolation transformer, the power line ground can be moved to where the oscilloscope's probe ground is attached.

[KungFuJosh]

Thanks!

[bhishmar]

That is how I do it. The idea is that with the isolation transformer, the power line ground can be moved to where the oscilloscope's probe ground is attached.

Would you please care to elaborate on the above  point?

I have read carefully both KungFuJosh query & ur reply.
But I am missing the point on "power ground being moved to DSO-ground"

[rstofer]

That is how I do it. The idea is that with the isolation transformer, the power line ground can be moved to where the oscilloscope's probe ground is attached.

Would you please care to elaborate on the above  point?

I have read carefully both KungFuJosh query & ur reply.
But I am missing the point on "power ground being moved to DSO-ground"

As soon as you clip the scope probe ground to some point in the circuit, that point is now earth ground by way of the BNC connector and the fact that the scope is grounded.  That is the way it's going to turn out whether intended or not.



Note that GFCI protection does not extend beyond the isolation transformer because it is looking for the current balance between the hot and neutral wires that only reach the primary of the transformer.

[David Hess]

That is how I do it. The idea is that with the isolation transformer, the power line ground can be moved to where the oscilloscope's probe ground is attached.

Would you please care to elaborate on the above  point?

I have read carefully both KungFuJosh query & ur reply.
But I am missing the point on "power ground being moved to DSO-ground"

With power from a three pin wall socket which includes ground, neutral, and hot, ground is attached to neutral at the circuit breaker box.  Any current through neutral, which should be all of it, raises the potential of neutral compared to ground so if any ground from the same circuit breaker box like the ground lead on the oscilloscope's probe is attached, it creates a ground loop and diverts current away from neutral which at best will corrupt the measurement and at worst break things.

The isolation transformer allows moving the ground point from the circuit breaker box to any point on the isolated side of the transformer with some limitations due to capacitance coupling across the transformer and capacitive coupling into the environment.  Now the ground lead on the oscilloscope's probe can be attached at that point relatively safely without creating a ground loop.

The ground connection on a differential probe would also be attached at that point to maximize common mode rejection.  Note that a differential probe has a ground connection whether it is brought out or not and it relies on that ground connection to control its common mode voltage.  If it is not brought out as a separate connection, then it is relying on the ground connection through the power cord.

This is why differential probes should not be used with oscilloscopes which have isolated inputs and for the same reason, they should not be used with isolation transformers unless a separate ground connection is made to the isolated output of the transformer somewhere.  Without the ground connection, the common mode voltage into the differential probe can drift all over the place.

[KungFuJosh]

With power from a three pin wall socket which includes ground, neutral, and hot, ground is attached to neutral at the circuit breaker box.  Any current through neutral, which should be all of it, raises the potential of neutral compared to ground so if any ground from the same circuit breaker box like the ground lead on the oscilloscope's probe is attached, it creates a ground loop and diverts current away from neutral which at best will corrupt the measurement and at worst break things.

The isolation transformer allows moving the ground point from the circuit breaker box to any point on the isolated side of the transformer with some limitations due to capacitance coupling across the transformer and capacitive coupling into the environment.  Now the ground lead on the oscilloscope's probe can be attached at that point relatively safely without creating a ground loop.

Your IT info sounds incomplete or assumptive/misleading unless I'm reading it wrong. In my case, my IT has an isolated floating neutral but still uses common ground (3 prong). My DUT chassis ground is still earthed through this.

Your description sounds like it would be more accurate with an IT and a non-grounded DUT (for example a 3 to 2 prong grounding adapter on the DUT plug).

[tautech]

With power from a three pin wall socket which includes ground, neutral, and hot, ground is attached to neutral at the circuit breaker box.  Any current through neutral, which should be all of it, raises the potential of neutral compared to ground so if any ground from the same circuit breaker box like the ground lead on the oscilloscope's probe is attached, it creates a ground loop and diverts current away from neutral which at best will corrupt the measurement and at worst break things.

The isolation transformer allows moving the ground point from the circuit breaker box to any point on the isolated side of the transformer with some limitations due to capacitance coupling across the transformer and capacitive coupling into the environment.  Now the ground lead on the oscilloscope's probe can be attached at that point relatively safely without creating a ground loop.

Your IT info sounds incomplete or assumptive/misleading unless I'm reading it wrong. In my case, my IT has an isolated floating neutral but still uses common ground (3 prong). My DUT chassis ground is still earthed through this.

Your description sounds like it would be more accurate with an IT and a non-grounded DUT (for example a 3 to 2 prong grounding adapter on the DUT plug).

Just confusion and seeing it from a different angle.
A recent thread might offer some additional clarity:
https://www.eevblog.com/forum/beginners/question-isolation-transformer-ground-floating-or-ground-shorted-to-neutral/

[KungFuJosh]

Just confusion and seeing it from a different angle.
A recent thread might offer some additional clarity:
https://www.eevblog.com/forum/beginners/question-isolation-transformer-ground-floating-or-ground-shorted-to-neutral/

I guess I'm a little confused by this:

The isolation transformer allows moving the ground point from the circuit breaker box to any point on the isolated side of the transformer

It's all still connected on the same ground. The only difference is the Neutral is separated from the ground. So what's the difference with connecting the ground?

The thing that confuses me is that it reminds me of a well-known "genius" who actually suggested using a 3 to 2 grounding adapter on an IT to connect regular probes anywhere in the circuit like you would with appropriate differential probes.

[David Hess]

Your IT info sounds incomplete or assumptive/misleading unless I'm reading it wrong. In my case, my IT has an isolated floating neutral but still uses common ground (3 prong). My DUT chassis ground is still earthed through this.

Your description sounds like it would be more accurate with an IT and a non-grounded DUT (for example a 3 to 2 prong grounding adapter on the DUT plug).

No, it still applies because for the same reason you cannot safely attach the oscilloscope probe's ground to the line side circuit anywhere, you cannot attach the power line ground either.  The ground connected to the chassis is not connected to neutral except of course back at the circuit breaker box.

The isolation transformer allows connecting the oscilloscope probe's ground to any part of the line side with some limitations duplicating the situation where neutral and ground are tied together at the circuit breaker box.  For safety however, the AC ground connection should also be attached to that point.

I guess I'm a little confused by this:

The isolation transformer allows moving the ground point from the circuit breaker box to any point on the isolated side of the transformer

It's all still connected on the same ground. The only difference is the Neutral is separated from the ground. So what's the difference with connecting the ground?

The difference is that with current flowing through neutral, neutral and ground at the device are at different potentials so there is no safe point to attach ground.

The thing that confuses me is that it reminds me of a well-known "genius" who actually suggested using a 3 to 2 grounding adapter on an IT to connect regular probes anywhere in the circuit like you would with appropriate differential probes.

That does not make any sense unless you mean using a 3 to 2 grounding adapter on the oscilloscope.  That does work for small or negative values of "work" and it even used to be recommended but it is definitely not safe for the oscilloscope or user.

[tautech]

Just confusion and seeing it from a different angle.
A recent thread might offer some additional clarity:
https://www.eevblog.com/forum/beginners/question-isolation-transformer-ground-floating-or-ground-shorted-to-neutral/

I guess I'm a little confused by this:

The isolation transformer allows moving the ground point from the circuit breaker box to any point on the isolated side of the transformer

It's all still connected on the same ground. The only difference is the Neutral is separated from the ground. So what's the difference with connecting the ground?

The thing that confuses me is that it reminds me of a well-known "genius" who actually suggested using a 3 to 2 grounding adapter on an IT to connect regular probes anywhere in the circuit like you would with appropriate differential probes.

I presume you mean on the isolated side, in which case there is no Neutral. There is only Live and Live WRT each other.
Isolation transformers isolate the current/voltage supply from mains ground and thereby reduce the hazard of phase (live) to neutral and/or ground shock risk.
Mains ground retained to the DUT is always preferable.

[KungFuJosh]

The thing that confuses me is that it reminds me of a well-known "genius" who actually suggested using a 3 to 2 grounding adapter on an IT to connect regular probes anywhere in the circuit like you would with appropriate differential probes.

That does not make any sense unless you mean using a 3 to 2 grounding adapter on the oscilloscope.  That does work for small or negative values of "work" and it even used to be recommended but it is definitely not safe for the oscilloscope or user.

Nope. The "genius" uses ungrounded DUTs on ITs. Exactly what I wrote. He literally has a video on youtube where he demonstrates doing that.

Sorry if the quotes on "genius" didn't convey my sarcasm strongly enough.

[KungFuJosh]

I presume you mean on the isolated side, in which case there is no Neutral. There is only Live and Live WRT each other.
Isolation transformers isolate the current/voltage supply from mains ground and thereby reduce the hazard of phase (live) to neutral and/or ground shock risk.
Mains ground retained to the DUT is always preferable.

Correct, and I agree. I don't like the guy's stupid video spreading stupidity, but I'm not trying to start a social media war, it would just make me look bad.

I get what you're saying, but when you wire a plug, or look at the IEC socket, it still says N on it.

[ogden]

When you connect scope ground clip to your device somewhere, you effectively ground that particular point of circuit to scope ground AND chassis of device. I do not find this safe, especially for your scope. Slightest error of yours or problem with equipment under test and least of your problems could be shopping for new scope.

You definitely shall use probe that is safe w/o isolation transformer.

[KungFuJosh]

When you connect scope ground clip to your device somewhere, you effectively ground that particular point of circuit to scope ground AND chassis of device. I do not find this safe, especially for your scope. Slightest error of yours or problem with equipment under test and least of your problems could be shopping for new scope.

I'm personally not an idiot (most of the time). I only connect the probe ground to the chassis ground.

You definitely shall use probe that is safe w/o isolation transformer.

I think something is lost in translation here. What do you mean?

[David Hess]

Nope. The "genius" uses ungrounded DUTs on ITs. Exactly what I wrote. He literally has a video on youtube where he demonstrates doing that.

Sorry if the quotes on "genius" didn't convey my sarcasm strongly enough.

Ah, well I do not know why one would bother doing that except in special circumstances.  One might for instance want to disconnect the chassis ground so that one does not inadvertently ground oneself or something else while poking about inside while the DUT is energized.  I might also want to use a different ground in place of the AC power ground if I thought electrostatic coupling was causing a problem.

When you connect scope ground clip to your device somewhere, you effectively ground that particular point of circuit to scope ground AND chassis of device. I do not find this safe, especially for your scope. Slightest error of yours or problem with equipment under test and least of your problems could be shopping for new scope.

You definitely shall use probe that is safe w/o isolation transformer.

That is why I mentioned attaching a real ground connection to the point where the oscilloscope's probe ground is connected.  At least then if there is a fault, some or most of the current will be diverted away from the oscilloscope.  It is not always feasible to avoid using a probe ground with differential probes although newer ones usually do not provide the option to do so even if you wanted to.

[KungFuJosh]

Ah, well I do not know why one would bother doing that except in special circumstances.  One might for instance want to disconnect the chassis ground so that one does not inadvertently ground oneself or something else while poking about inside while the DUT is energized.  I might also want to use a different ground in place of the AC power ground if I thought electrostatic coupling was causing a problem.

Well, in my case and his, we're talking about tube guitar amplifiers. There's always a metal chassis, and whether or not it's grounded is kinda a big deal. I know people can get away with doing that, but it's...so...stupid.

That is why I mentioned attaching a real ground connection to the point where the oscilloscope's probe ground is connected.  At least then if there is a fault, some or most of the current will be diverted away from the oscilloscope.  It is not always feasible to avoid using a probe ground with differential probes although newer ones usually do not provide the option to do so even if you wanted to.

My differential scope is powered by USB on the oscilloscope. Should I not do that, or is it cool?

[ogden]

You definitely shall use probe that is safe w/o isolation transformer.

I think something is lost in translation here. What do you mean?

Yes, it's late here. Ignore that. If every voltage you measure in tube amp is chassis-referenced, then all you need is hi-voltage 1:100 probe, or two. Original probe is 150V rated in 1:1 and 300V rated in 1:10, I would not use such in 400..500VDC (or whatever voltages you have there) tube amp, especially knowing how sloppy that switch is.

[David Hess]

That is why I mentioned attaching a real ground connection to the point where the oscilloscope's probe ground is connected.  At least then if there is a fault, some or most of the current will be diverted away from the oscilloscope.  It is not always feasible to avoid using a probe ground with differential probes although newer ones usually do not provide the option to do so even if you wanted to.

My differential scope is powered by USB on the oscilloscope. Should I not do that, or is it cool?

It does not matter either way.  Just use it as designed.

[KungFuJosh]

Yes, it's late here. Ignore that. If every voltage you measure in tube amp is chassis-referenced, then all you need is hi-voltage 1:100 probe, or two. Original probe is 150V rated in 1:1 and 300V rated in 1:10, I would not use such in 400..500VDC (or whatever voltages you have there) tube amp, especially knowing how sloppy that switch is.

I've changed my setup, but I'm well covered. I returned the DS1054Z and bought a Siglent SDS1104X-E instead. I have Pico TA131 250MHz probes, which are rated 600V in 10X, and I have a BK PR2000B 100X probe that's rated 2000V. I did also buy a differential probe, but that's 1300V at 500X, which isn't always useful.

[David Hess]

I've changed my setup, but I'm well covered. I returned the DS1054Z and bought a Siglent SDS1104X-E instead. I have Pico TA131 250MHz probes, which are rated 600V in 10X, and I have a BK PR2000B 100X probe that's rated 2000V. I did also buy a differential probe, but that's 1300V at 500X, which isn't always useful.

Keep in mind that the voltage rating of those probes only applies when the oscilloscope is set to DC coupling.  If AC coupling is used, then the full input voltage is applied to the oscilloscope input which is probably only rated to 400 volts.

[KungFuJosh]

Keep in mind that the voltage rating of those probes only applies when the oscilloscope is set to DC coupling.  If AC coupling is used, then the full input voltage is applied to the oscilloscope input which is probably only rated to 400 volts.

Yup, 400. One of the benefits over the Rigol which only could handle 300. But even when I eventually figure out what the benefit of changing the coupling type is then it's all good because most of my circuits don't go above 350.

[David Hess]

Keep in mind that the voltage rating of those probes only applies when the oscilloscope is set to DC coupling.  If AC coupling is used, then the full input voltage is applied to the oscilloscope input which is probably only rated to 400 volts.

Yup, 400. One of the benefits over the Rigol which only could handle 300. But even when I eventually figure out what the benefit of changing the coupling type is then it's all good because most of my circuits don't go above 350.

You might want to use your x10 or x100 probe to measure low ripples voltages which would require AC coupling to work at a higher vertical sensitivity.

They make, or used to make, special x10 and x100 probes which included an internal shunt allowing them to be used with DC or AC coupling without limitations.  x1000 probes always include this shunt.

[bhishmar]

Part of the earlier confusion here is created by  non-standard or arbitrary  terminology  used in the posts here, by different poster's I think.
Same is the case with the other post referred by TauTech.

So let me define/ re-define  some symbols , & put forward a topology here, as I understood from earlier posts, to aid in any further discussion (if any).
==================================================================
IT    :  Isolation Transformer
DUT  :  Device under Test.

E1      :  Earth-Pin (3rd pin)  of Mains Power receptacle
L     : Phase (Live) point of Power receptacle
N    : Neutral point of Power receptacle.

L1    :  Output-Line-1 of   IT Output Power  receptacle.
L2    :  Output-Line-2 of   IT Output Power  receptacle.
E2    :  Earth-Pin (3rd pin) of  IT-output receptacle.

** I am not claiming above is standard terminology.  Just that some terminology aids in streamlining discussion & help in better understanding.
==================================================================

Standard wiring configuration inside Isolation Transformer (IT)

L    &  N   wired to primary winding of IT    (obviously)
L1  & L2  wired to secondary winding of IT (obviously)
E1 & E2   both wired to Chassis of IT. (so E1 & E2 are shorted inside IT)
==================================================

Typical Wiring outside IT   (Topology):-
DUT  is powered from a Bench power supply /adapter  which is powered from L1 & L2 of  IT.
E2 is connected to  the chassis of Bench PowerSupply. (or adapter if it has a 3rd pin Power-connector), thru power-cord

Bench Power-Supply chassis may or may not be connected to its DC-supply output -ve terminal (& hence DUT Circuit Gnd) - Tester's option.

If using a DSO directly powered from mains (not thru IT), its BNC-Gnd and any DUT-point  touched by black-alligator clip will be at Mains-Earth (E1) point.
If DSO is powered thru IT, then BNC-Gnd will be same as E2.
But since E1 & E2 are shorted, ultimately BNC-Gnd (DSO-Gnd) will be E1, in both cases.
==================================================

I am assuming in earlier posts  "Power-Gnd"   meant  E1,
& "DSO-Gnd" meant BNC-Gnd.

So "Power-Gnd" (E1)  &  DSO-Gnd (BNC-Gnd) will always be shorted & same, irrespective of  DSO is powered directly or thru IT !!

[bhishmar]

The isolation transformer allows moving the ground point from the circuit breaker box to any point on the isolated side of the transformer with some limitations due to capacitance coupling across the transformer and capacitive coupling into the environment.  Now the ground lead on the oscilloscope's probe can be attached at that point relatively safely without creating a ground loop.

Would you care to explain the above in the context of above terminology & topology?
I understand  there are 3 separate issues to contend with.
1.  Ground loop creation by  DSO-Gnd & associated noise introduction.
2.  Some Ground continuity path thru a hi-impedance for differential-probe's (internal ground), to limit common-mode voltage at differential amplifier inputs.
3. Saftety