Author Topic: My first scope is arriving but my home (old) doesn't have mains grounding  (Read 1451 times)

0 Members and 2 Guests are viewing this topic.

Offline Fergo

  • Contributor
  • Posts: 7
  • Country: br
I am not an engineer but I am an electrician. Some people have weird ideas in this thread. An RCD or GFCI is a very good idea since if current that flows out through L1 and doesn't return on L2 wil cause it to trip. This is the danger with having no group d at the out let since you can become a path to ground just by touching the metal enclosure of the device. The danger in not having a grounding means at your outlet is that you could have damage or a fault in your device or dut and energize something that shouldn't be, if supplied from a grounded outlet this would cause your breaker or fuse to trip and deenergize the circuit. Since you don't have that the next best thing is to detect when current is flowing outside of the intended circuit and deenergize at that point. This is what an RCD/GFCI will do.

Sent from my SM-G965W using Tapatalk

Thanks!
But from what I've read in other replies, you do need a ground wire at least in the distribution box for the RCD do work. I don't have that :(

One way or another. I couldn't be happier. The scope arrived and I've been playing and getting used to it almost non stop on my free time:



It's so cool (and yes, the Arduino is connected to my laptop running on batteries). Definitely worth it's price, considering I managed to get it new with almost 50% off.

KR,
Fergo
 
The following users thanked this post: ogden, Electro Detective

Offline Electro Detective

  • Super Contributor
  • ***
  • Posts: 2041
  • Country: au
I am not an engineer but I am an electrician. Some people have weird ideas in this thread. An RCD or GFCI is a very good idea since if current that flows out through L1 and doesn't return on L2 wil cause it to trip. This is the danger with having no group d at the out let since you can become a path to ground just by touching the metal enclosure of the device. The danger in not having a grounding means at your outlet is that you could have damage or a fault in your device or dut and energize something that shouldn't be, if supplied from a grounded outlet this would cause your breaker or fuse to trip and deenergize the circuit. Since you don't have that the next best thing is to detect when current is flowing outside of the intended circuit and deenergize at that point. This is what an RCD/GFCI will do.

Sent from my SM-G965W using Tapatalk

OP Fergo states in the post that there is no Earthing,

better to not blow money on a RCD that won't work (although the TEST button might  :horse:)

and buy up on PPE instead,
and a fast fuse and MCB combo.

Personally, in that situation, I would prefer two unearthed lines coming from a properly fused isolation transformer humming nearby,
rather than two unearthed lines coming directly from the street transformer via power station,
and relying on an aged sticky mechanism 16 to 30 amp breaker to trip at the mains board = good luck and happy barbeque!   :scared:

 
« Last Edit: August 14, 2019, 07:50:01 pm by Electro Detective »
 
The following users thanked this post: Fergo

Offline Ian.M

  • Super Contributor
  • ***
  • Posts: 7819
No ground at outlets does *NOT* mean the building is free of grounds.  Its likely that any metal conduit and pipes go to ground, however their ground resistance is almost certainly too high to safely trip a breaker, and the integrity of any ground they might appear to provide is suspect. 

A RCB (GFCI) requires a ground referenced supply to operate, but it is extremely unlikely that the supply isn't ground referenced.  In fact its virtually certain the neutral is grounded at the supply company's neighbourhood distribution transformer.  The only non-ground referenced (floating) supplies you are ever likely to encounter are from isolation transformers or portable generators etc. operated without a ground rod.  RCDs don't do anything useful if fed from a floating supply.

Using a RCD (GFCI) does therefore increase safety in the O.P's situation as it will trip on any hazardous leakage to ground, even if the ground path is too high resistance to trip the existing circuit breaker or fuse.
 
The following users thanked this post: Fergo

Offline kamtar

  • Contributor
  • Posts: 20
which can provide some risks as the USB negative is theoretically tied to earth (theoretically because my computer is not grounded either, as the sockets don't have the ground pin).

just a little bit of off-topic: most notebooks aren't grounded unless you connect an external monitor or something like that. There is only some minor leakage current as with every switching power supply.
 
The following users thanked this post: Fergo

Offline Ian.M

  • Super Contributor
  • ***
  • Posts: 7819
That depends on the notebook or rather its PSU 'brick'.  If its got a two pin mains connector you can be certain its *NOT* grounded.  However most OEM notebook PSUs with a three pin connector have the output 0V grounded.

I say OEM, because there is a flourishing trade in fake replacement PSUs assembled in the far east from a clone of the OEM case moulding and whatever PSU board salvaged from eWaste, of near enough the same output voltage they  can find to cram in there and those often omit any connection from the mains connector ground pin.  Also beware of fake power leads with three pin connectors on both ends but only wired with two wire cable.

Anyway, its easy enough to check - with the PSU unplugged, check continuity from the ground pin to the contacts of the DC output connector.  If they aren't easy to probe, plug the DC connector in and check continuity from the ground pin of the mains connector (still unplugged) to the shell of a USB socket or similar on the notebook.
 
The following users thanked this post: Fergo

Offline Electro Detective

  • Super Contributor
  • ***
  • Posts: 2041
  • Country: au
As stated earlier, ignoring the onboard TEST button for a moment, unless an RCD can be verified with an external tester to simulate a fault,
and the RCD/GFCI performs reliably with consistency as per it's time/current specs (and not just a couple of 'lucky' trips  :clap:

then OP Fergo is wasting money on a false sense of security buying into an RCD in a situation where it may/might/won't/ work, or all three together  :horse:
 
As it is, even with an RCD in place in a properly grounded system, if I'm playing 'risky business' with electronicals
there is my own portable fast trip 10>15ma RCD, or RCD/MCB combo close by,
so Murphy can stay on holidays and not bother me with cheap travel offers to hot places  >:D

It may be two RCDs to reset if a larger fault trips the main RCD too, but hey, still better value than post zap hospital therapy
or zapping the relatives with a funeral bill

FWIW, you only have to spend maybe one weekend to assemble all the generic bits needed get the safety thing right ONCE,
to save you and your gear MANY MANY times.

It's actually fun doing it, testing the lot just to see and know what the outcomes and limitations are,
and realize after a few trips and sort of controlled BANGS!  :-[ or pending snafus  that it's not so predictable as you thought to make a system idiot proof. 

Cheap fast blow inline fuses, even no name knockoff/fakes? (pre-test a couple first!) rated just above the expected current draw,
will save you LOTS of money on blunders that are guaranteed to come knocking.

I've got an assortment from 100ma up to 10amps, and the best $20 investment I spent years ago.
If it saves even one cheap multimeter or project or repair, you just got $20 back and the rest  :-+

An isolation transformer (or three..  :palm:) with switchable modes of earth/ground isolation is a great tool, but not for those unfamiliar with those modes.

Still, a proper wired 'medical' style isolation transformer is a good bet for 'some' isolation and interference/noise from the mains,
even though the earth/ground and neutral output are bonded internally.
I use those whenever I can as they will only pump out whatever power their limit is, and some expensive types will regulate and current limit on an overload or dead short.   


But all that stuff is sort of useless unless you've simple PPE eyewear/face shield, gloves and rubber soles,

A bucket of sand, fire blanket and fire extinguisher wouldn't hurt either and won't break your pocket   

I keep all that stuff in one flaky carton box, which saves on looking around for the bits, and doubles as a traveller too

Stay safe and best of luck to OP with the new scope  :-+





 
 
The following users thanked this post: Fergo

Offline Mechatrommer

  • Super Contributor
  • ***
  • Posts: 8939
  • Country: my
  • reassessing directives...
RCD/GFCI needs an Earth/ground reference to work,
no...

an imbalance between the Live 1 and Live 2 conductors (Active and Neutral) during a fault won't do anything after passing through the internal sensing coil
yes if you buy one hung low brand...

As stated earlier, ignoring the onboard TEST button for a moment, unless an RCD can be verified with an external tester to simulate a fault,
and the RCD/GFCI performs reliably with consistency as per it's time/current specs (and not just a couple of 'lucky' trips  :clap:)
have you encountered a properly brand working and installed RCD (2 wires LN in LN out), that wont trip during imbalance? regardless of ground line existence or not? afaik it will trip even if current lost in the air..

read under the "Purpose and operation" https://en.wikipedia.org/wiki/Residual-current_device
Quote
Operation does not require a fault current to return through the earth wire in the installation; the trip will operate just as well if the return path is through plumbing, contact with the ground or any other current path. Automatic disconnection and a measure of shock protection is therefore still provided even if the earth wiring of the installation is damaged or incomplete.
the only time you get a shock when current through you from live back to neutral, so no imbalance inside RCD. hence dont use both hand..
if something can select, how cant it be intelligent? if something is intelligent, how cant it exist?
 
The following users thanked this post: Ian.M

Offline forrestc

  • Supporter
  • ****
  • Posts: 467
  • Country: us
RCD/GFCI needs an Earth/ground reference to work,
no...

I'll second this.

In the USA in particular it is perfectly acceptable to replace a 2 wire outlet with a 3 wire GFCI.   The principle here is that a 3 wire GFCI provides more protection than a 2 wire outlet would provide.   

The only protection you lose in the US system by doing this is not providing a path for leakage currents/shorts to find their way to ground.   With a good ground, a loose wire touching the grounded case will result in a trip of the circuit breaker or GFCI.    Without the ground, the path will be potentially through a human to the ground, a path which will cause the GFCI to trip.

I don't know what the OP's national electrical system looks like, so I can't be 100% sure this is the same in their country.   But I can attest that a GFCI absolutely does not require a ground to work.
 
The following users thanked this post: SilverSolder, Ian.M

Offline Ian.M

  • Super Contributor
  • ***
  • Posts: 7819
I don't know what the OP's national electrical system looks like, so I can't be 100% sure this is the same in their country.   But I can attest that a GFCI absolutely does not require a ground to work.
There is a small error in your statement: A GFCI does *NOT* require a ground at its point of installation, but as I said before, it does require its supply to be ground referenced.

Nearly all utility company mains supplies in the world are ground referenced. 
Exception: some parts of Norway and Finland, where due to low ground conductivity (typically due to massive igneous rock), older homes may be on an IT system (i.e. floating) supply.

A GFCI wont work on a fully floating supply as it works by using a balanced current transformer to detect any imbalance between the current in the two (or more) live supply conductors to determine if there is a significant leakage current to ground, and, considering a two conductor GFCI directly fed from an isolating transformer secondary that is not connected to ground, unless the secondary itself is faulty, there is no way there can be an imbalance as the current in the wires to the two ends of the winding must be equal and opposite, no matter what fault is introduced on the load side of the GFCI, so it will never trip on any load fault.

A GFCI's built-in test button introduces an imbalance by connecting a resistor diagonally across the GFCI terminals, from one conductor on the supply side to the opposite one on the load side, so the test resistor current only flows in one conductor passing through the GFCI.   That will trip just that GFCI even if its fed from a floating supply that its otherwise incapable of tripping on.   If the test button introduced a resistor to ground, it could potentially trip all GFCIs upstream of it and which one(s) actually tripped would depend on their relative speed and trip time vs current curves.
« Last Edit: August 15, 2019, 04:53:15 pm by Ian.M »
 

Offline Mechatrommer

  • Super Contributor
  • ***
  • Posts: 8939
  • Country: my
  • reassessing directives...
A GFCI's built-in test button introduces an imbalance by connecting a resistor diagonally across the GFCI terminals, from one conductor on the supply side to the opposite one on the load side, so the test resistor current only flows in one conductor passing through the GFCI.   That will trip just that GFCI even if its fed from a floating supply that its otherwise incapable of tripping on.
but if one touch either wire downstream of the floating supply, the current will leak to wherever the person is standing to, as long as it is not returned to the other line, it will create current imbalance in the GFCI.
if something can select, how cant it be intelligent? if something is intelligent, how cant it exist?
 

Offline Electro Detective

  • Super Contributor
  • ***
  • Posts: 2041
  • Country: au
Believe what you will, flog the math and have faith,
but if an RCD/GFCI does not respond reliably and repeatedly to an external simulated fault test, and at 180 degrees,
good luck.

Then there's fault tripping and SMPS noise issues to contend with.

If the power system in OPs case is suss, playing with non isolated input oscilloscope and other single ended test gear,

he/she needs to be 100% sure a failsafe system in place works 

i.e. you rely on an RCD/GFCI to bail you out, it works for a while (apparently... :-//)  and one day it either fails to work, or was never going to work, 
and or the system earthing develops issues over time, or some electricity maintenance friday helper twat that works for beer disconnects it,
and the RCD faither may end up in the ground, with the 'unfortunate event' passed off as 'misfortune' or 'bad luck' 

An RCD is still better than nothing,
but nothing with extreme caution is good too, especially when one solo RCD won't work when you need it to

FWIW, even the best RCD will just sit there and look on  :popcorn: while the owner cooks on a live chassis with no working earth ground



« Last Edit: August 16, 2019, 08:26:39 am by Electro Detective »
 

Offline Mechatrommer

  • Super Contributor
  • ***
  • Posts: 8939
  • Country: my
  • reassessing directives...
how about having an RCD, and be extremely careful? well i can think one safest foolproof way, be in a cave away from electricity. my wife is wiser to stay away from all this shit. remember, all this taboo is already there since the OP dwell that apartment, regardless if he buy a scope or not. unearthed refrigerator, unearthed tv, unearthed hair drier unearthed water heater unearthed everything, or the unreliable RCD thereof.
if something can select, how cant it be intelligent? if something is intelligent, how cant it exist?
 

Online Jeroen3

  • Super Contributor
  • ***
  • Posts: 3196
  • Country: nl
  • Embedded Engineer
    • jeroen3.nl
A GFCI's built-in test button introduces an imbalance by connecting a resistor diagonally across the GFCI terminals, from one conductor on the supply side to the opposite one on the load side, so the test resistor current only flows in one conductor passing through the GFCI.   That will trip just that GFCI even if its fed from a floating supply that its otherwise incapable of tripping on.
but if one touch either wire downstream of the floating supply, the current will leak to wherever the person is standing to, as long as it is not returned to the other line, it will create current imbalance in the GFCI.
No significant current will flow aside from the capacitive leakage in the transformer.

...
If the test button introduced a resistor to ground, it could potentially trip all GFCIs upstream of it and which one(s) actually tripped would depend on their relative speed and trip time vs current curves.
Considering some RCD's run amps of current trough their test resistor when their rating is only 30mA, yes, that would be a problem.
 

Online ArthurDent

  • Frequent Contributor
  • **
  • Posts: 792
  • Country: us
There is only one correct solution to solving this problem and that is to use a proper ground (earth) from the building’s A.C. power source that meets the current local code. Where this doesn’t appear to be immediately possible, anything done will be a kludge to one degree or another. What I would consider doing is constructing your own 3 wire A.C. supply completely isolated from the building’s mains supply and require no changes or additional connections to the building’s A.C. supply.  Solar would work but a complete solar system with storage and conversion is totally out of the question and impractical.

Here is a possible solution for you to discuss and critique that is a little different than some of the other suggestions above. What I think might work, if you’re just going to power your own little lab, is to buy an isolation transformer with enough capacity to power everything you could connect to the mains supply and then wire the secondary as you would a properly grounded 3 wire A.C. line from the power company. At the transformer secondary one lead would be the high side line and the other secondary lead, right at the transformer, would be the low and also have a ground or PE wire which would be a non-current carrying conductor, just like a properly grounded 3 wire A.C. supply. These wires would be treated just like a regular 3 wire A.C. line with a correctly rated breaker, GFCI/RCD, and outlets on your bench. Because there will be some capacitance coupling you might want to consider three about 330K ohm resistors in series from the core and the shield between the primary and secondary windings to the low primary lead to drain any small leakage - see drawing.

This should work o.k. especially if your lab is located where there are no earth referenced conductors coming in from the outside like phone lines, tv cable, or water pipes. The idea is to create what amounts to a space station or clean room type environment that is galvanically isolated from the outside so there is no possible accidental connection to outside earth yet you still have a local 3 wire system for your bench that will allow safety circuits to be used normally.
 
The following users thanked this post: Electro Detective

Offline Fergo

  • Contributor
  • Posts: 7
  • Country: br
I don't know what the OP's national electrical system looks like, so I can't be 100% sure this is the same in their country.   But I can attest that a GFCI absolutely does not require a ground to work.

Here's is a schematic of a distribution box in accordance to our code (or as the author said so...), which is mandatory since 1997. My building is from the 70s, so it's nothing like this:


In this case, it's a 2-phase system (220V). The red device is an SPD. The big one in the middle is an RCD/GFCI. The other ones are regular breakers (2-phase input).

This is an English speaking forum, but if eventually someone is interested, here is our current building code for low voltage installations:
http://universidadeniltonlins.com.br/wp-content/uploads/2019/04/NBR-5410.pdf


« Last Edit: August 16, 2019, 05:58:46 am by Fergo »
 

Online Jeroen3

  • Super Contributor
  • ***
  • Posts: 3196
  • Country: nl
  • Embedded Engineer
    • jeroen3.nl
Does a GFCI also interrupt on broken ground or neutral? Since you all have a ground wire connected to it.
I've never seen a RCD in a european store with a ground wire or terminal on it.

Is that an extra feature required by code? Since broken ground or neutral is very rare here with all underground wiring.
 

Offline Mechatrommer

  • Super Contributor
  • ***
  • Posts: 8939
  • Country: my
  • reassessing directives...
it should as per wiki i linked and from logic thinking of how it is constructed, any current imbalance will create non zero magnetic field and open the elecromagnetic coil circuit inside it, it alone doesnt required ground to work, ground wiring is imho only for double fold safety for current return path safely to earth in case wiring fault downstream such as live wire broken inside appliance and touching exterior body. with earth wiring, current flow to ground, gfci tripped, if no earth wiring, no current flow, gfci not tripped, the exterior body just float on mains potential waiting for somebody to touch to complete circuit for current to flow back to earth. without earth wiring, human body is the only mean for current to flow, but if there is earth wiring, both will share the current, either way, as long as there is current lost (non zero net) inside gfci, it will tripped, such as when a person touch bare live line (no appliances needed), i accidentally tripped gfci this way several time, never it tripped due to appliances fault (live touching metallic body) because i think products design are robust enough or i got lucky to not have shabby one hung low items in the house. not sure why is this difficult to be understood that it has to be explained several times. otoh, broken wiring is another matter, gfci should not be blamed for that.

ps: why gnd wire is connected to neutral uptream gfci? to create low impedance path to ground localized to that house/area in case neutral line got raised in the middle by lightning, short on the pole or whatever (i've seen corona short somewhere, real magnificent!). by right neutral line is earthed back in the dist transformer, if not, the contractor who installed it should be blamed (at least in our country). if someone installed ground wire to neutral downstream the gcfi, thats why gfci probably not tripped in case electric shock, the current will flow from human to earth back to gfci downstream back to gfci, no current difference, net is zero, gfci will not tripped.
if something can select, how cant it be intelligent? if something is intelligent, how cant it exist?
 

Online Jeroen3

  • Super Contributor
  • ***
  • Posts: 3196
  • Country: nl
  • Embedded Engineer
    • jeroen3.nl
Your GFCI are dual throw and short downstream to ground? I know this is how it can work on high voltage, but for low voltage is a bit weird...
 

Offline Mechatrommer

  • Super Contributor
  • ***
  • Posts: 8939
  • Country: my
  • reassessing directives...
Your GFCI are dual throw and short downstream to ground?
no! dont do that i already explained that just before :palm: it should be upstream ie before the GFCI to your lab. if you look at reply #21, following mains supply from right to left, earth wiring is connected to neutral before the GFCI, ie after the isolator (the big CB, high current trip for all distribution smaller CBs, leftmost white component in Fergo diagram reply #39) but i dont check inside of my switchbox, whether it should be after or before isolator. i think its better after the isolator since everything (LNE) will be disconnected when isolator is tripped. what is certain, is NE connection should be before GFCI (RCCB). someone may check their certified switchbox to see at which point E and N is connected together. GFCI only for mains supply 110-220V, for non lethal low voltage you dont need any of that.
« Last Edit: August 16, 2019, 10:03:37 pm by Mechatrommer »
if something can select, how cant it be intelligent? if something is intelligent, how cant it exist?
 

Online Jeroen3

  • Super Contributor
  • ***
  • Posts: 3196
  • Country: nl
  • Embedded Engineer
    • jeroen3.nl
I asked why an GFCI's has a ground terminal. Not why the distribution network has a neutral-earth bond. That is obvious.
I recently developed an automated tester for IEC 61009/62423 RCD's, and a ground terminal on an GFCI is not something I have seen before.
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 12712
  • Country: gb
  • Hero999
Yes, an RCD is a good idea.

If it's not possible to fix the wiring in the building, then your best bet is to use an RCD and connect all metal objects (metal water, heating pipes etc.) and conductors which are supposed to be earthed together with equipotential bonding to ensure it all stays at the same voltage.
 

Offline Mechatrommer

  • Super Contributor
  • ***
  • Posts: 8939
  • Country: my
  • reassessing directives...
the my GFCI only 2 input 2 output. live and neutral input, live and neutral output (single phase). ground and neutral tied somewhere inside switch box, that is another story from GFCI.
edit: ok you made me work. the neutral line is not connected to my ground, i was talking out of an arse just now. but neutral to ground is only 1.7Vrms ;D
if something can select, how cant it be intelligent? if something is intelligent, how cant it exist?
 

Offline Electro Detective

  • Super Contributor
  • ***
  • Posts: 2041
  • Country: au

It won't matter to OP Fergo who is right, wrong, or somewhere in between,  :-//
if he/she gets zapped, takes out equipment, any suggested wiring hacks affect other building dwellers,
or all three scenarios, making for REAL news on the idiot box (TV) 

There is no guaranteed 'protection magic' buying and installing an RCD/GFCI, unless tested under real world conditions,
and tested regularly. I've replaced and tested/pre-tested enough of them to have a bit of a clue..  \$\Omega\$

They can fault out in too many ways and display no symptoms of failure or pending failure,
unless tested accordingly to verify that the trip current and trip times are up to spec and repeatable,
and notes taken including ambient temperature. 
Every six months or 12 months is a good bet, and if the RCD/GFCI is suss, then swap it out with a pre-tested QUALITY unit made by ThreeHungLow, rebadged for some former decent manufacturer.      

AFAIK the main marketing idea is that if the 'victim to be' touches something metallic that's connected to the earth/ground bonded neutral, and the active/live wire together,  :o
the RCD will protest  >:( :rant:  and no longer supply electricity,

saving the already battling household an unwanted funeral bill,
and the hassle of flogging the NLR hobby gear on Ebay to get some money back to eat.


FWIW RCD/GFCI can work nicely on the output of an isolation transformer, and not affect electrical codes, insurance policies etc

..if done correctly, according to where you are, be it 120v, 240v and what the earthing arrangement is 

If not done correctly, RCD will either not work or get cooked internally on the first TEST button flick, and never work again    :horse:



 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf