Non-Earth Grounded, Oscilloscope Observation [video]

[BrooklynNerd]

I'm wondering if anyone could help me understand what is going on here.

Here's the video: http://youtu.be/S9u1J4L3Cy8

In a nutshell, power in my apartment building is not up to code. Most of my wall outlets are not earth grounded. I've plugged a new oscilloscope into one of those outlets and the results are interesting.

• The probe laying on a table picks up a 60hz signal, yes- I know it's picking up the mains
• The the signal is amplified when I grab the probe w/o touching the metal prong on the probe .
• When I touch the metal prong, Vmax jumps even more to ~1.70

When I earth ground the plug to my heater things change:

• Resting probe on the table, or me holding the without touching the metal no longer picks up 60hz
• When I touch the metal probe, I get a much lower 60hz Vmax ~180mV

My question is, what exactly is happening when my wall outlet is not earth grounded? Why is there so much extra noise? I thought earth ground was a safety feature. Does it also help discharge unintended capacitance?

Is it normal to see 1.2 Vrms going through my body?

Is it safe to earth ground it by attaching the third prong to my heater?

Thanks everyone.. and Thanks Dave for your awesome videos!  I'm a CS guy and am trying to take the leap from micro-controllers to more serious EE.

[c4757p]

The filter capacitors in the power supply are connected to ground, so when the ground pin isn't connected, they apply voltage to the chassis and the shields on the input connectors (possibly quite a lot, up to half the mains voltage). Because the oscilloscope reads (V_tip - V_shield), the relatively large voltage now present on the shield becomes a term in that expression, and unless V_tip follows it closely, you're going to see noise.

It's not really unsafe to ground it like that for signal purposes only IMHO (others may disagree on this topic, we've had lots of arguments about it...), but you must not trust it for safety, because you don't know how good of a ground it is.

You really need to get proper grounding installed for your lab. As I'm sure many people will point out, grounding isn't absolutely essential for safety, but grounding items that are designed to be grounded is.

Keep in mind that when the chassis isn't grounded, that voltage comes through the PSU filter capacitors, so the only thing separating you from the equivalent of sticking your finger in a light socket is a little capacitor. Do you trust it? Granted, they're probably good ones in that brand-new scope, but that doesn't necessarily mean they're safe.

[IanB]

A possible way to imagine the earth ground on an oscilloscope is like a big heavy object that damps out vibrations. If you wanted a very stable table free of vibrations you would put a big heavy metal slab on it (this is the kind of thing people do when they are making holograms where everything needs a very stable platform to avoid messing up the image). The earth ground on an oscilloscope is kind of similar--it is an electrical equivalent of a big heavy metal slab, anchoring your oscilloscope to the stable voltage of the earth so that it can then measure the tiny voltage fluctuations you are interested in without electrical noise getting in the way.

[BrooklynNerd]

Thanks all, that is helpful. This apartment building is 100+ years old and sometimes when I run into these sort of problems, I am never certain how what I am experiencing compares to the norm.

It also serves as a learning experience.

[Paul Price]

BrooklynNerd, See the previous topic below for more info on this topic.

https://www.eevblog.com/forum/beginners/is-it-safe-to-use-a-bench-power-supply-not-grounded/

https://www.eevblog.com/forum/beginners/new-lab-no-grounded-outlets/

-------------------------------------------------------

C4757p,
There is a great difference between working with ungrounded equipment such as an oscilloscope, and sticking your finger into a light socket.

If you stick you finger in a light socket you are exposing you finger to both wires of the AC mains and  you can be almost sure a good shock is possible...at least across your finger, but maybe not across you body, but it could be uncomfortable. If you are standing in a bathtub of water while doing this, that could be a different story.

If you touch the metal chassis of an ungrounded piece of say, an oscilloscope, you are possibly only in contact with a quite small capacitance's tiny AC current reactive leakage at the line frequency but the current has nowhere to flow and there is no shock unless you have other earth-grounded metal objects you are also in physical contact with.

You could have the possibility of experiencing a mild shock from that tiny capacitor only if you can reverse a possible unpolorized socket plug and now can get the tiny current through the capacitor, but only, again if you are otherwise in contact with some grounded metal object at the same time.

If you are living in an USA household that has two-wire outlets, one wire is neutral and is connected to earth ground at the power entry point breaker box of your house and does not pose a danger, if that small capacitor in the oscilloscope is connected to the neutral side of the power line.

[c4757p]

I am obviously referring to the case of capacitor failure. I understand reactance.

[amyk]

That is almost certainly going to be a Class Y rated capacitor which is designed to fail open.

[madires]

BrooklynNerd, See the previous topic below for more info on this topic.

https://www.eevblog.com/forum/beginners/is-it-safe-to-use-a-bench-power-supply-not-grounded/

https://www.eevblog.com/forum/beginners/new-lab-no-grounded-outlets/

Don't rely on a single RCD, if one is installed! Add more layers of protection! A proper ground adds more protection and reduces some measurement problems.

[c4757p]

That is almost certainly going to be a Class Y rated capacitor which is designed to fail open.

Indeed it is. Failing short is still sonething that one should at least consider, though, in part based on the quality of the device. Hence "do you trust it?". Personally, I would, but it is something to be aware of.

If you are living in an USA household that has two-wire outlets, one wire is neutral and is connected to earth ground at the power entry point breaker box of your house and does not pose a danger, if that small capacitor in the oscilloscope is connected to the neutral side of the power line.

Assuming some half-witt didnt reverse the L-N when wiring. Neutral is earth grounded at the utility or pole I think, not the individual residence. I could be wrong its been a while since I've looked at residential wiring codes and power distrubution.

At the breaker board, usually. And Paul, it's typically one capacitor on line and one on neutral, not just a single on neutral...

[Paul Price]

C4757p, BTW thanks for the tip on the laser printer, it was the one that solved the problem.

But I must say it depends on the circuit designer, most AC Line power supplies I've seen (I've seen hundreds) seem to use just one cap to ground, some very much fewer use one from both sides of the line, like refrigerators and electric stoves, dishwashers, heavy white stuff makers like to do that.

In all my many years of troubleshooting, I've never seen one of these rather funny looking  capacitors (with all kinds of strange various country safety symbols printed on them) short out.  Doesn't mean it can't ever-ever happen.

[c4757p]

These very common line inlet EMI filters frequently have one capacitor across the line, plus one capacitor each from line and neutral to ground. My Rigol DS1052D, with presumably the same PSU board as his, uses the exact same circuit on the board itself. I don't know about "most" power supplies, but it's very common particularly in oscilloscope power supplies.

I agree, I've never seen them short out. As I said, I would trust them - but I wouldn't advise someone else to. Rather, I'd advise him to do the research and come to the same conclusion I have. If somebody tells me "don't do that, it's dangerous", I listen unless I have good reason to think he's full of shit, but if somebody tells me "oh, it's fine, go ahead", I verify first. Anything that should be an insulator protecting you from mains or high voltage shorting out is always something you should think about until you know it's not going to.

[BrooklynNerd]

Thanks again, this is all very good info. I should have searched the forum better and will definitely check out those other posts.

[leb120]

For buildings where adding a ground would be too expensive or not possible a GFCI is a very good alternative.
How they work: http://www.2d2c.com/gfci.php

If you own the building and feel comfortable replacing a receptacle you can get a GFCI that will replace your existing plug.
In a 100 year old building you may want to avoid this though; the wiring may have cloth insulation and may disintegrate when you touch it.

Just turn off your breaker first.

The other option is to get a GFCI protected power strip. Only items plugged into the strip will be protected, you can plug other power strip into this one and they will be protected as well.
https://www.google.com/search?q=gfci+power+strip&tbm=shop

You could still tie the ground wire to your water or steam pipe (avoid gas piping) before of after the GFCI.

GFCIs (a.k.a. GFI) look at the current coming in on the hot or neutral and compare that to the current leaving on the other wire. If there is an imbalance (because some of the current is flowing through you) of just a few milliamps it will open (disengage) both prongs of the outlet within a few tens of milliseconds. Most of the time you won't even feel the shock it just protected you from.

Even a relatively poor "bonding" path back to the panel (through water piping) can still trip a GFCI meaning if a any equipment does have a ground fault (such as capacitors shorting) the GFCI will trip and protect you.

[Spunky]

If the heater connection removes the AC wave from your scope then there must be an earth which means a GFCI will work, but only if the earth is either bonded to ground (earth rod), or there is a ground to CPC connection near to you from a nearby building, enough to carry over 5mA current.

If not then you would be on an 'earth island' where a phase to earth fault simply raises the potential of the ground and no current flows (no GFCI trip), at which point of course both phases are equally live or neutral to you (take your pick), and neither can harm you unless you touch both, but dont test this though in case your ground is connected to earth. Of course if you have multiple class1 appliances then this is hazardous as 2 faults could see 2 adjacent appliances energised to each phase, waiting for you to close the circuit with your hands. And then the GFCI doesn't save you because the current killing you is matched on both phases 

So, a GFCI can't hurt, but it may not save you unless you bond the supply earth to ground.

As for the scope, yes earth it to a heater. In the absence of any other earth why not?

[Tepe]

So, a GFCI can't hurt, but it may not save you unless you bond the supply earth to ground.

Could you please write to the authority responsible for the thousands of pages of wiring code in my country and inform them that they have been mistaken for decades?

[alm]

Do they seriously protect floating mains circuits with GFCIs in your country? Can you name one situation when a GFCI connected to a floating supply would trip? A GFCI detects leakage current, i.e. current flowing through one of the two wires, and then back via some path outside the GFCI (resulting in imbalance). If both wires are floating (no connection between supply earth and ground), then there is no alternate path. If you are connected to earth and touch one of the wires (say live) then not much will happen because the circuit will just float so live is at earth potential. No current will flow, and nothing will trip. If you now touch the other wire (the two class I appliances in Spunky's example), then you are connected between live and phase, and the GFCI will happily supply the full 20A indefinitely.

So in this situation a GFCI is either useless and the situation is safe, or the GFCI is useless and you are being fried. Illustrating the importance of an earth-referenced mains for GFCI operation.

[IanB]

So, a GFCI can't hurt, but it may not save you unless you bond the supply earth to ground.

Could you please write to the authority responsible for the thousands of pages of wiring code in my country and inform them that they have been mistaken for decades?

I read that Denmark uses a TT supply arrangement, where the "T" stands for "terra", or earth. The generator or transformer neutral is connected to earth at the supply, and the consumer has a local independent earth. So the electrical supply in Denmark is certainly connected to earth, and that is why an RCD will work.

[Tepe]

Do they seriously protect floating mains circuits with GFCIs in your country?

I was being facetious. Of course the mains is grounded somewhere. Even if the OP should happen to be in the IT using parts of Norway, it will be grounded somewhere which makes Spunky's remark "So, a GFCI can't hurt, but it may not save you unless you bond the supply earth to ground" miss the point. Mains will not be floating freely unless you add an isolation transformer or you have serious installation flaws. The former is under your direct control and it is a fair guess that the latter is far less likely than you accidentally connecting yourself as a load between live and neutral - something neither grounded equipment nor CFGIs will protect you from.

[Tepe]

I read that Denmark uses a TT supply arrangement, where the "T" stands for "terra", or earth. A the generator or transformer neutral is connected to earth at the supply, and the consumer has a local independent earth. So the electrical supply in Denmark is certainly connected to earth, and that is why an RCD will work.

Correctamundo. TN-C-S is also seen.

[Paul Moir]

Here it's acceptable to replace an ungrounded receptacle with one that's GFCI protected.  That is to say you can use a 3 prong outlet on a ungrounded circuit so long as it's GFCI protected.  Ungrounded receptacles are impossible to find anymore, and rewiring is expensive and therefore requiring it prevents improving safety.  The utility and your panel do the ground bonding.

But please don't quote me; you know, some guy on the internet, since at the very least it's been a long time since I've looked at the Canadian Electrical Code.

[Spunky]

Well, I've seen plenty of these RCD problems on installations which have no supply earth (usually temporary installations in the UK), and it can be a hell of a job sometimes to get a good enough earth to trip the RCD. So it's not like I haven't been there and done it. A shiny new 4ft earth spike driven in deep with an SDS drill usually works, but not always if the ground is poor. There needs to be enough current returning to it's origin.

My point was a TT earthed system might trip an RCD or GFCI, and then again it might well not, and seriously they often don't. TT is antiquated, not up to modern safety standards, and there's a good reason why most of the world regulated it out of existence for new homes.

If TT and IT are good enough for us all then the rest of europe has wasted it's time and money, we can forget about earth loop impedance testing and Ze values and relax, but it's not the case. I suspect Denmarks reluctance to adopt modern standards are more about economics than welfare.

Admittedly GFCI is a lot lower trip current than RCD. Which would help, but doesn't negate my point. GFCI/RCD might work without a decent earth and is better than not having one.

[Monkeh]

Well, I've seen plenty of these RCD problems on installations which have no supply earth (usually temporary installations in the UK), and it can be a hell of a job sometimes to get a good enough earth to trip the RCD. So it's not like I haven't been there and done it. A shiny new 4ft earth spike driven in deep with an SDS drill usually works, but not always if the ground is poor. There needs to be enough current returning to it's origin.

And if the earth is so poor it won't trip, enough current to cause damage (to equipment or persons) can't flow..

My point was a TT earthed system might trip an RCD or GFCI, and then again it might well not, and seriously they often don't.

In which case the shock risk is minimal, too!

TT is antiquated, not up to modern safety standards, and there's a good reason why most of the world regulated it out of existence for new homes.

TT is a perfectly acceptable method and is fully capable of meeting modern safety standards with an appropriately designed installation.

Admittedly GFCI is a lot lower trip current than RCD. Which would help, but doesn't negate my point. GFCI/RCD might work without a decent earth and is better than not having one.

A GFCI and an RCD are the same thing. They can trip at 5mA, 10mA, 15mA, 30mA, 100mA, 300mA, 1A, instantly, with a time delay, etc.

[Tepe]

If TT and IT are good enough for us all then the rest of europe has wasted it's time and money, we can forget about earth loop impedance testing and Ze values and relax, but it's not the case. I suspect Denmarks reluctance to adopt modern standards are more about economics than welfare.

I think you suspect wrong. TT is alive and kicking outside the English speaking world. It is possibly even the most widespread system and it does have certain advantages.

[Spunky]

There are certainly some advantages to TT, like the reduced risk of fire, and no risk of a broken supply earth. TNS and TNCS have the advantage that they have better safety against electrocution, better safety against faulty RCD/GFCI, and fast tripping of MCBs in a phase to earth fault. Choice depends on the supply infrastructure, which as it upgrades usually results in a shift away from TT and IT systems.

My point about driving an earth rod applies specifically where there are multiple class 1 appliances, to prevent 2 phase faults being dangerous. Not needed if everyone is running laptops and TVs (class2), but really important if they are running fridges and cookers.

RCD and GFCI are not the same thing. RCD is passive, the current imbalance in the coil itself trips it, but a GFCI uses a sensor circuit connected to the coil, hence they usually work on lower fault currents, and depend on both phases reaching the GFCI.

[Monkeh]

RCD and GFCI are not the same thing. RCD is passive, the current imbalance in the coil itself trips it, but a GFCI uses a sensor circuit connected to the coil, hence they usually work on lower fault currents, and depend on both phases reaching the GFCI.

Would you care to provide some evidence of this? Because they really are the same fundamental device, and an RCD can be configured for virtually any trip current you like.