[doctorcilantro]

Fitting the subwoofer amp died today, rather, started to fail like the last one. I think the DC was killing it. Luckily I have a 3rd backup and it is quiet - when they die they start to spew static like a bad tube.

So far, dishwasher is on, and today at lunch and tonight, I have heard no buzzing on the speakers or my PS Audio P3 AC regenerator.

Amazing how much DC must have been on the line? But I only measure .08v DC...not sure if voltmeter can read something like this accurately though....

[jea48]

My original statement was:

A GFCI does not require a Safety Ground (EGC) reference! The GFCI only examines the current in the Send and Return wires, if these are not equal (5ma) then the unit trips.

Just how is that statement incorrect?

***************************************************
from your first link:

 Don't let the name confuse you — these devices will operate on a circuit that does not have an equipment-grounding conductor.

Hi Speedskater,

There is nothing wrong with your above statement. As I said in my post the GFCI does not need an (EGC) to operate properly. We are in agreement.

The GFCI does not need or use a external reference.  It is only aware of the current in the Send and Return wires. As long as these two currents are about equal the GFCI is happy.  In any case, current is not interested in going to the ground or into the dirt,  it is only interested in getting back ti it's source which is that big power company transformer outside.

I do not agree with you here though. If a GFCI is connected to a 2 wire Grounded AC power system it will operate as designed. It will not operate as designed if connected to an Isolated AC power System.
Jim

[jea48]

Thank you Occam for the drawing.

To Jea48

As for the service neutral conductor / earth /equipment enclosure/ equipment ground be bonded at only one location at the main electrical service of the building we do not know that it is not.

Please refer to the photo from the OP, it clearly shows the neutral conductor being bonded to discoloured copper cables, which eventually end up on the metal conduits. The neutral in then bonded to earth, not in a manner that's normal, surely.

Another point of using GFCI/RCD is that if the earth conductor is broken on say a toaster or an appliance outside, if someone touches the live portion of the appliance while say touching a tap/faucet, the GFCI will trip.

In your first test, if you wired the device correctly, the GFCI should have tripped. If it didn't, it's great you found a dud. The RCD is supposed to trip since it artificially creates an imbalance of current around the CT, rather than through it.

Thank you Occam for the drawing.

Agree! I suggest you look closely at the "Test" circuit. What do you see?


Please refer to the photo from the OP, it clearly shows the neutral conductor being bonded to discoloured copper cables, which eventually end up on the metal conduits. The neutral in then bonded to earth, not in a manner that's normal, surely.

Sure is normal....  Look closely at the latest pictures posted by the OP. Just to the right of the feeder conduit you will see a conduit with a bare copper wire exiting the conduit. That bare copper wire is the Grounding Electrode Conductor that bonds the service neutral conductor to the Earth Grounding Electrode.


Another point of using GFCI/RCD is that if the earth conductor is broken on say a toaster or an appliance outside, if someone touches the live portion of the appliance while say touching a tap/faucet, the GFCI will trip.

CORRECT! Why? Because the " GFCI/RCD" is connected to a Grounded AC Power System. The "someone" completes a closed circuit and current flows.


 In your first test, if you wired the device correctly, the GFCI should have tripped. If it didn't, it's great you found a dud. The RCD is supposed to trip since it artificially creates an imbalance of current around the CT, rather than through it.

True. I should have double checked that one before posting. I edited my post this AM. MY BAD for not double checking.
Jim



   

[jea48]

Agreed, GFCI can be used on floating supplies. To Jea48, how then are delta/delta transformers protected, since there is no common star point?

The GFCI will not operate as designed on a floating power supply. It will not protect squat.

To Jea48, how then are delta/delta transformers protected, since there is no common star point?


Back in the day when a 3PH delta power system was commonly used in industrial facilities one leg of the delta was intentionally connected to earth making it the Grounded Conductor. Example a Grounded 3PH 3 wire 440V AC Power system.  This system was quite common in non-hazardous facilities.

In a small industrial facility it was not uncommon to see a 3PH 4 wire delta power system. Example a 3PH 4 wire 230/115V Grounded AC Power system.

For full power 3 transformers are used. One split phase single phase transformer for 230/115V, and two smaller single transformers for 3PH delta power. The center leg of the split phase single phase transformer is intentionally connected to earth making it the Neutral, the Grounded Conductor.
With this Grounded 3Ph 4 wire power system a "Wild Leg" is created. It was common to see a 3PH 4 wire main service panel board used for power distribution. The Wild Leg is installed on "B" phase of the panel board and is identified by orange color.

Voltage measurements from phase to phase? 230V nominal.
From phase "A" to ground? 115V nominal.
From phase "B", (Wild Leg) to ground? Can vary anywhere from around 208V to 277V. (This leg is not used for power from phase to neutral.)
Voltage measurement from "C" phase to ground? 115V nominal.

Last but not least an un-grounded 3 phase 3 wire delta power system used in a hazardous location facility.
Good example is a soybean mill/oil extraction facility.
Back in the day this system used a crude simple Line Isolation Monitor circuit to warn the qualified personnel one of the floating phases had went to ground. IF, a ground fault condition occurred it was the job of the maintenance electricians to find the fault and correct the fault as soon as possible. LOL, Imagine the Prep Building with hundreds of motors trying to find where the fault might be. The Prep building was never shut down to find the ground fault problem. The Extraction Building was another matter. Hexane is very explosive. http://en.wikipedia.org/wiki/Hexane
Any ground fault detection in the Extraction Building Plant operators shut down the operation in the building and the fault was found and repaired.

Back to the Prep Building. Bean dust can be explosive but in the old plant “downtime” meant lost profit. Find the problem if you can just don't stop production. Otherwise, wait for the next scheduled shutdown to find the problem.
So the old Prep building continued running with one leg of the 3 phase 440V grounded by the ground fault condition. Everything was fine beans flowing..... Yep until another motor or motor branch circuit wire went to ground on another phase. You want to guess what happens next?
Say the first ground fault occurred in the pecker head of a 100 HP motor.  Motor vibration had caused a poorly insulated "A" phase makeup to bare and come in contact with the grounded motor pecker head.  The later second ground fault occurred in a small 3/4 HP motor where the ground fault occurred on  "C" phase.  Want to bet which Motor starter do you think lost the battle?


Bottom line, NEC Code prohibits the use of a Isolated Power System, (Floating the secondary of an ISO xfmr), in a residential occupancy. A GFCI will not protect a person from electrical shock in any way what so ever when connected to a floating power system. Line to line or line to ground.

On a 2 wire grounded AC power system a GFCI will trip open if a ground fault occurs from a Line to earth grounded object. If the current flow is greater than 5mA the GFCI will trip open.
A GFCI will not protect a person from Line to Line lethal electrical shock.
Jim   

 

[One and a half]

It appears that Kirchoff's law is applied in only certain circumstances Jea48?

The core balance CT in the GFCI on one leg will pick up the current to earth as well as the load current, but invariably the leakage current through a different path, to earth, will be missed by the CT.
The resultant differential is enough to trip the GFCI, earthed system or floating.

High voltage transmission systems use IT or TT systems for protection of three/six/twelve wire systems. Final earthing of the neutral in most countries is at the consumer premises level with a double wound transformer. Interesting story of the factory, sounds like NEC methods.

[jea48]

FWIW

 One and a half,

 I am still waiting for your answer on how the "Test" circuit works in the diagram Occam provided in his post.




 

[One and a half]

One and a half,

 I am still waiting for your answer on how the "Test" circuit works in the diagram Occam provided in his post.

By pushing the test button, this allows a current to flow through the CT since the resistor is connected across the two incoming lines (not marked in the drawing as to the input).
The resistor current returns to the upstream connection bypassing the CT. Since the current drawn by the resistor is not cancelled out by the return wire from the resistor, the CT and electronics measure the difference in current I load + I resistor. If the I resistor current is more than the device's sensitivity, the breaker will trip.

Various GFCI are designed to measure either sine waves, or pulsed currents for example from rectifiers/SMPS. The device needs to be correctly specified for the intended function. More reading: https://www.hqs.sbt.siemens.com/gip/general/dlc/data/assets/hq/manual/beta_all_technology-primer-rcpd_tpi_en.pdf

The circuit does not require any of the main conductors (Hot & neutral) to be grounded, since grounding is not relevant for displacement currents. So long as the load current remains symmetrical, the GFCI will hold.

Typical current paths. The lady's "fault current" is only seen by one leg of the CT, the GFCI will trip.






For a 3phase network, here's a schematic showing the terminals.




You are correct, nothing will prevent a line to line load from operating a GFCI, if that load were a person. If the over current device works, then this may be the saving grace and applies equally to systems protected or not protected by GFCI. And the point of which is?

[jea48]

By pushing the test button, this allows a current to flow through the CT since the resistor is connected across the two incoming lines (not marked in the drawing as to the input).
The resistor current returns to the upstream connection bypassing the CT. Since the current drawn by the resistor is not cancelled out by the return wire from the resistor, the CT and electronics measure the difference in current I load + I resistor. If the I resistor current is more than the device's sensitivity, the breaker will trip.

Various GFCI are designed to measure either sine waves, or pulsed currents for example from rectifiers/SMPS. The device needs to be correctly specified for the intended function. More reading: https://www.hqs.sbt.siemens.com/gip/general/dlc/data/assets/hq/manual/beta_all_technology-primer-rcpd_tpi_en.pdf

The circuit does not require any of the main conductors (Hot & neutral) to be grounded, since grounding is not relevant for displacement currents. So long as the load current remains symmetrical, the GFCI will hold.

Typical current paths. The lady's "fault current" is only seen by one leg of the CT, the GFCI will trip.






For a 3phase network, here's a schematic showing the terminals.




You are correct, nothing will prevent a line to line load from operating a GFCI, if that load were a person. If the over current device works, then this may be the saving grace and applies equally to systems protected or not protected by GFCI. And the point of which is?

We are making progress, YES? ....

Now if we can only get you to understand why a GFCI will not operate as designed on an Isolated Power System.

Thanks for posting the diagrams. Pictures and diagrams are worth a thousand words. Note in the diagram with the lady the ground fault current path. Follow the current path of the ground fault. Power source L3 >>> through the lady >>> Earth ground >>> back to the source the Grounded Neutral Conductor. "Houston we have a lift off"...... A completed circuit.

Now if you would be so kind as to show me a diagram or explain the completed ground fault circuit path of a floating secondary of an ISO transformer to the main grounding electrode system of the premises the primary of the ISO transformer is fed from.


http://www.schneider-electric.us/sites/us/en/support/faq/faq_main.page?page=content&country=US&lang=EN&id=FA177591&redirect=true
Question: Can a GFCI be used on an isolated power system?
Answer: GFCI devices should never be connected to an Isolated Power System. GFCIs will not operate properly and can result in unwanted ground connection in the Isolated Power System.


http://products.schneider-electric.us/technical-library/?event=detail&oid=09008926803f2737&cat=0b008926800ae93f
Scroll down page, click on view.

Next please go to page 12.

Quote.
Per NEC® 100, “A Ground-Fault Circuit Interrupter (GFCI) is a device
intended for the protection of personnel that functions to de-energize a
circuit or portion thereof within an established period of time when a current
to ground exceeds the values established for a Class A device. A Class (A)
ground-fault circuit interrupters trip when the current to ground has a value
in the range of 4–6 mA.” Unlike an IPS, a GFCI can only be employed on
solidly grounded power distribution systems.


Next please scroll down to page 16.
Quote.
An IPS is intended to operate as a strictly ungrounded power system. GFCI can only be utilized on solidly grounded power
systems. The benefits of utilizing an IPS are not only the minimizing of arcing, sparking, and fault currents that may ignite an
explosion or fire. IPSs can also reduce shock hazards from the inadvertent contact with a phase conductor and can allow for
the continuation of electrical appliance or machine if a phase conductors of the branch circuit were to experience an actual
phase-to-ground fault condition. If a GFCI were to sense an actual phase-to-ground fault condition, then the power in the
branch circuit would be interrupted.


Quote]:
You are correct, nothing will prevent a line to line load from operating a GFCI, if that load were a person. If the over current device works, then this may be the saving grace and applies equally to systems protected or not protected by GFCI. And the point of which is?
One and a half

 And the point of which is?

Look at it this way, a typical home has a single phase 240V (120/120) service. If the GFCI circuit is connected to the 'A' leg/pole (aka phase) and it's Hot 120V wire leaks to the 'B' leg's Hot 120V. There will be leakage current (after all these two Hots are 240 Volts apart) so the GFCI will trip.

We are not talking about  ICs or speaker cables here. We are talking about electrical power that can kill. I have only stuck with this thread in the hope of convincing the laymen reader that a GFCI will not protect a person from a lethal line to line shock if no other part of the person’s body is in contact with a grounded object.

Second a GFCI  WILL NOT  provide any protection whats so ever when connected to a floating secondary of an ISO transformer.

Just curious, what country do you live in? 

 


 

[One and a half]

Yes, you are right, I am wrong.

What  if the impedance between the two grounds are high, such as in a TT or IT system. A core balance relay, GFCI/RCD will still pick up the leaks to ground, since the return path is not symmetrical. The Schneider documents apply to medical regulations. This document from ABB explains more to the world market, than for a specific regulation in the one country.

http://www05.abb.com/global/scot/scot209.nsf/veritydisplay/c4e584f06cc6c4fbc1257ad800496193/$file/2CSC420004B0201_RCDs%20EN.pdf

For IT systems, core balance will still work, but will need a secondary system to discriminate the type of fault. For TT systems where the supply ground is a some impedance to the consumer's ground, the core balance systems works (eg. Japan).

For the isolation transformer for audio purposes, grounding one leg of the secondary has the advantage to localise and create a lower impedance path to earth for the audio equipment as far as noise is concerned. It's just not that simple to join all earths at the one point where the audio equipment is spread out with extension cables, plug boards and the like.

In this case a GFCI better yet RCBO will work (since you can't see how it can work any other way) and will work better than a fuse or breaker due limited amount of short circuit current available from the transformer.

[doctorcilantro]

Hate the resurrect this thread....

Big wind storm. Lost power 3 times. Tons of trees down.

The buzz has returned from being 100% absent.

My plan:

To move the circuit to the third remaining leg, since testing with current and last used OFFICE leg (formerly AUDIO leg) result in the buzz on the P3 - you can hear it buzzing across the room.

[doctorcilantro]

Anyone think this is not a DC current issue?

I received an interesting email from Blue Circle about my noise problem. Seem to imply they think it is a grounding problem not a DC problem.

 To summarize...

    a buzzing started coming out my speakers, especially when the dishwasher was on added a P3 and the P3 torroid buzzes loudly, especially when the dishwasher was on
    The buzzing would also occur at random times day or night
    I moved the audio circuit to another leg and the problem vanished.
    After a huge wind storm with power outages the problem returned
    Also, our larger water-heater was replaced soon after the storm

Thoughts?

[Speedskater]

We are talking about electrical power that can kill. I have only stuck with this thread in the hope of convincing the laymen reader that a GFCI will not protect a person from a lethal line to line shock if no other part of the person’s body is in contact with a grounded object.

Second a GFCI  WILL NOT  provide any protection whats so ever when connected to a floating secondary of an ISO transformer.

Regardless of how many out of context or simplified examples you quote, this is incorrect!

A GFCI device only asks one question:
Is the amount of current on the Neutral the same amount as the Hot?  If not - it trips. End of story, it does not know or care about what is causing that current difference.

[Speedskater]

Perhaps my point would be easier to grasp if for the word 'ground' in the above articles we substitute the concept of:
"An accessible  pathway for fault current to flow back to it's source and it's source is the power company Neutral"

********************************
'ground' is not a destination, it is a path.

*******************************
And once again, a GFCI device does not need to be connected to a Safety Ground (EGC) to operate correctly.

[jea48]

Regardless of how many out of context or simplified examples you quote, this is incorrect!

A GFCI device only asks one question:
Is the amount of current on the Neutral the same amount as the Hot?  If not - it trips. End of story, it does not know or care about what is causing that current difference.

LOL, If the secondary of an ISO transformer is allowed to float please explain how the current in either of the two hot legs could be different.  And, LOL, there is no neutral...... No reference to ground from either hot ungrounded hot leg what so ever.....

LOL just couldn't resist....
Jim

[jea48]

Perhaps my point would be easier to grasp if for the word 'ground' in the above articles we substitute the concept of:
"An accessible  pathway for fault current to flow back to it's source and it's source is the power company Neutral"

********************************
'ground' is not a destination, it is a path.

*******************************
And once again, a GFCI device does not need to be connected to a Safety Ground (EGC) to operate correctly.

I didn't say a GFI receptacle needed to be connected to a safety equipment ground to operate as designed when connected to an AC grounded power system .
A GFI receptacle will not operate as designed if connected to the secondary of an ISO transformer if the secondary is floating.
You are an EE, a simple bench test would prove your theory wrong. 
You might want to reread the thread.

[Speedskater]

OK, here is the difference.  You are looking at it from a theoretical system viewpoint. That is with an isolation transformer just where is the fault or leakage current going to go?  But the GFCI does not look outside itself, all it can see is the current on the Hot & Neutral wires and if these two currents aren't the same but opposite then it trips.  Now getting back to the real world, that ISO transformer might be good today but what about tomorrow?

[jea48]

OK, here is the difference.  You are looking at it from a theoretical system viewpoint. That is with an isolation transformer just where is the fault or leakage current going to go?  But the GFCI does not look outside itself, all it can see is the current on the Hot & Neutral wires and if these two currents aren't the same but opposite then it trips.  Now getting back to the real world, that ISO transformer might be good today but what about tomorrow?

There is no neutral on a 2 wire ungrounded floating AC power system, only 2 hot ungrounded conductors.  I am still waiting for you to explain how there could be a current difference on either of the two hot ungrounded conductors. For there to be a differential/imbalance of current between the 2 hot ungrounded conductors a leakage of current on one hot ungrounded conductor would need to exist to ???, to cause an imbalance. Or one hot ungrounded conductor would have to have more or less leakage to ??? for an imbalance to exist.

Again a simple bench test would disprove your theory.
* Place an ISO transformer on a table.
* Connect the primary winding to a suitable AC source that will deliver 120V out on the secondary winding.
* Connect a GFI receptacle to the two secondary output leads of the ISO transformer.
* plug a load into the receptacle.

Now please explain your theory of how or where a current imbalance could possibly take place for the sensing coil, of the GFI receptacle, to sense any imbalance what so ever anywhere in the completed 2 wire 120V circuit.


 Now getting back to the real world, that ISO transformer might be good today but what about tomorrow?

Not sure what you mean.....

Jim

[Speedskater]

Now please explain your theory of how or where a current imbalance could possibly take place for the sensing coil, of the GFI receptacle, to sense any imbalance
If there is no current imbalance there is no safety problem.

 Now getting back to the real world, that ISO transformer might be good today but what about tomorrow?
Not sure what you mean.....

What I mean is when everything is operating correctly then everything is OK, but what happens if the transformer insulation fails? or a wire from another circuit somehow comes in contact with the ISO circuit.

GFCI's are only important when things go wrong.

[jea48]

Now please explain your theory of how or where a current imbalance could possibly take place for the sensing coil, of the GFI receptacle, to sense any imbalance
If there is no current imbalance there is no safety problem.

 Now getting back to the real world, that ISO transformer might be good today but what about tomorrow?
Not sure what you mean.....

What I mean is when everything is operating correctly then everything is OK, but what happens if the transformer insulation fails? or a wire from another circuit somehow comes in contact with the ISO circuit.

GFCI's are only important when things go wrong.

I asked,
Now please explain your theory of how or where a current imbalance could possibly take place for the sensing coil, of the GFI receptacle, to sense any imbalance

Your answer,
If there is no current imbalance there is no safety problem.

Please give an example of an event that could cause an imbalance.....

What I mean is when everything is operating correctly then everything is OK, but what happens if the transformer insulation fails? or a wire from another circuit somehow comes in contact with the ISO circuit.

When properly connected to a grounded AC power system,
A GFCI does not sense any leakage whatsoever upstream of the device. The GFCI only detects a problem downstream of the device.
Jim

[doctorcilantro]

And.....it's F'ing back.

With a vengeance. I have traced it to the ground. I can reduce it 90% but unplug my sub and amp which both connect to the amps speaker terminals.

With my system off I still get some buzz at speakers, tonight, it's really buzzing.

The PS Audio P3 is in front of basic Cullen outlet, and I have a custom DC blocker built for the amp and my sub.

Tonight I ran the P3 into kitchen and den outlets, on different circuits, and it still howls.

Here is the P3 torroid from about 12" away:

[Ps Audio P3 Torroid](https://clyp.it/n35sbgcy)

Here is me standing at the doorbell transformer, I never heard this rise and fall over the existing hum before, then it's me trudging back to my speakers where you can the buzz with my system powered down:

https://clyp.it/3ftfo5kt

I think we must have asymmetrical 3phase power here,, this is some serious DC on the line. NOTHING STOPS IT!!!!!!!!!