The house is circa 1950's single level on a concrete slab in California (lightning is very rare) and has existing romex two-wire residential branch circuits ran through the attic. Is it unreasonable to simply install GFCI receptacles vs pulling a ground to all receptacles (which would pretty much mean upgrading the entire electrical service.) GFCI's should be code compliant in such a case and would afford the occupants a greater degree of safety than the existing system. However, I'm concerned that much of our modern electronics probably need a ground to operate as designed. Am I missing anything here?
Electronics on Ungrounded Residential Circuits
- Thread starter Ripter
- Start date
- Status
- Location
- Chapel Hill, NC
- Occupation
- Retired Electrical Contractor
Article 406.4D(2) allows you to do this. You don't have to necessarily have a gfci in every receptacle as you can feed thru and use a 3 wire receptacle but you must use the stickers to say that there is no ground
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
151207-1435 EST
Ripter:
What is sensitive electronic equipment? What is this equipment to be protected from? How does an EGC protect the equipment? How would a GFCI protect the equipment? What is the response time of a GFCI? Does a GFCI protect you from getting a shock? If you are on a ladder, get shocked from a GFCI circuit, and fall off can a GFCI protect you from being killed?
.
Ripter:
What is sensitive electronic equipment? What is this equipment to be protected from? How does an EGC protect the equipment? How would a GFCI protect the equipment? What is the response time of a GFCI? Does a GFCI protect you from getting a shock? If you are on a ladder, get shocked from a GFCI circuit, and fall off can a GFCI protect you from being killed?
.
- Location
- Placerville, CA, USA
- Occupation
- Retired PV System Designer
There could be problems if you run signal wires with "grounded" reference conductor between components which are not separately grounded and you do not provide isolation.
Various undesired couplings to 60Hz or transient could result in high current in the signal ground.
But the best solution is to provide isolation instead of adding EGCs.
Various undesired couplings to 60Hz or transient could result in high current in the signal ground.
But the best solution is to provide isolation instead of adding EGCs.
- Location
- Massachusetts
I agree that 406.4D(2) allows GFCIs in place of EGCs.
However, in my opinion 250.114 prohibits plugging technology equipment or much of anything into those receptacles without EGCs.
However, in my opinion 250.114 prohibits plugging technology equipment or much of anything into those receptacles without EGCs.
qcroanoke
Sometimes I don't know if I'm the boxer or the bag
- Location
- Roanoke, VA.
- Occupation
- Sorta retired........
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
151207-2001 EST
Ripter:
The questions I ask you were intended to stimulate thought. I don't plan to answer them, but to make some observations.
What I say below is not intended to imply what I would or wouldn't do, or suggest to be done.
The intent of your original post seemed to be directed to:
An EGC is a normally non-current carrying conductor connected (bonded) to the neutral only at the main panel, and in turn is grounded (earthed) by an approved means. The purpose of an EGC is to provide a means to connect all exposed conductive surfaces on any electrical equipment connected to the EGC to earth.
Under normal conditions this should mean that the electrical potential between any electrical devices connected to the EGC and other conductive objects (cement floors, plumbing pipes, gas pipes, etc) is near zero volts and a human touching both should not get a shock.
It also means that leakage current within an EGC connected device should not appreciably raise the potential of the EGC at the leakage point relative to earthed conductive objects. If the EGC connected equipment is connected to a GFCI protected circuit, then this also means that the leakage current can not exceed the threshold limit of the GFCI for very long without tripping the GFCI.
If the circuit is not GFCI protected (a GFCI takes time to respond so it provides no protection for a short time), then a dead short between the hot conductor in the electrical equipment to the EGC in that equipment can raise the EGC potential at the shorting point relative to earth to approximately 1/2 the voltage of the hot conductor (multiply the RMS value by 1.414 to get the peak voltage) until the circuit breaker for the circuit opens. In some cases this will destroy electronic equipment. In particular un-isolated RS232, 422, 485, and USB circuits. Ethernet endpoints are usually isolated by a transformer.
Different pieces of electronic equipment connected together via non-isolated paths should have all the AC cords originate from the same receptacle. Not complete protection, but good protection from shorts outside of the said equipment.
.
Ripter:
The questions I ask you were intended to stimulate thought. I don't plan to answer them, but to make some observations.
What I say below is not intended to imply what I would or wouldn't do, or suggest to be done.
The intent of your original post seemed to be directed to:
What is ground? Is it just an EGC? Do most laptop computers, cellphone chargers, TVs, hairdryers, electric drills with double insulation, and many other devices with internal electronics have a two or three prong AC power cord plug.
An EGC is a normally non-current carrying conductor connected (bonded) to the neutral only at the main panel, and in turn is grounded (earthed) by an approved means. The purpose of an EGC is to provide a means to connect all exposed conductive surfaces on any electrical equipment connected to the EGC to earth.
Under normal conditions this should mean that the electrical potential between any electrical devices connected to the EGC and other conductive objects (cement floors, plumbing pipes, gas pipes, etc) is near zero volts and a human touching both should not get a shock.
It also means that leakage current within an EGC connected device should not appreciably raise the potential of the EGC at the leakage point relative to earthed conductive objects. If the EGC connected equipment is connected to a GFCI protected circuit, then this also means that the leakage current can not exceed the threshold limit of the GFCI for very long without tripping the GFCI.
If the circuit is not GFCI protected (a GFCI takes time to respond so it provides no protection for a short time), then a dead short between the hot conductor in the electrical equipment to the EGC in that equipment can raise the EGC potential at the shorting point relative to earth to approximately 1/2 the voltage of the hot conductor (multiply the RMS value by 1.414 to get the peak voltage) until the circuit breaker for the circuit opens. In some cases this will destroy electronic equipment. In particular un-isolated RS232, 422, 485, and USB circuits. Ethernet endpoints are usually isolated by a transformer.
Different pieces of electronic equipment connected together via non-isolated paths should have all the AC cords originate from the same receptacle. Not complete protection, but good protection from shorts outside of the said equipment.
.
Last edited:
Thank you Iwire, not exactly the answer that I was hoping for but thoughtful and well documented. Much appreciated.
Sorry GoldDigger (...Placerville...nice
you lost me a bit on the signal wire part but it sounds like we have something of a consensus that leaving the ungrounded system in place is less than ideal. Now comes the cost/benefit analysis...Thanks for the clarification Gar, I thought you were just being sarcastic and unhelpful in your first response.
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
151208-2121 EST
Ripter:
You really need to define what is "sensitive electronic equipment" that might be used in a home, and why operationally no EGC would present a problem. I grew up and lived in and worked in non-EGC environments for at least 30 years. Amplifier hum and you reversed the plug. An EGC might have automatically eliminated this, but an EGC connection might have created other noise problems.
Provide an example of equipment and its operation (not safety reasons) that would benefit from being connected to an EGC, and why.
There have been machine tool manufacturers that claim that noise problems can be solved by a "good ground" at the machine. No definition of what a "good ground" is, and/or why it will solve the problem. Some CNC servicemen (probably suggested by the CNC manufaturer) recomend disconnecting the CNC machine from the EGC, and instead use a local ground rod for grounding to solve noise problems. This is wrong, violates the code, and is not safe. Usually this relates to RS232 communication problems. There is more to this, but until you provide more information I don't have a direction to follow.
I do think you need the GFCI for people and equipment safety reasons in your two wire system.
If you have electronic equipment with a two prong plug, then an ECG is of no value unless the EGC is used in some external fashion.
.
Ripter:
You really need to define what is "sensitive electronic equipment" that might be used in a home, and why operationally no EGC would present a problem. I grew up and lived in and worked in non-EGC environments for at least 30 years. Amplifier hum and you reversed the plug. An EGC might have automatically eliminated this, but an EGC connection might have created other noise problems.
Provide an example of equipment and its operation (not safety reasons) that would benefit from being connected to an EGC, and why.
There have been machine tool manufacturers that claim that noise problems can be solved by a "good ground" at the machine. No definition of what a "good ground" is, and/or why it will solve the problem. Some CNC servicemen (probably suggested by the CNC manufaturer) recomend disconnecting the CNC machine from the EGC, and instead use a local ground rod for grounding to solve noise problems. This is wrong, violates the code, and is not safe. Usually this relates to RS232 communication problems. There is more to this, but until you provide more information I don't have a direction to follow.
I do think you need the GFCI for people and equipment safety reasons in your two wire system.
If you have electronic equipment with a two prong plug, then an ECG is of no value unless the EGC is used in some external fashion.
.
JFletcher
Senior Member
- Location
- Williamsburg, VA
As was already covered, you can use GFCI outlets in lieu of a grounding wire. If your house is wired anything like mine, you will need more GFCI than usual as many outlets are wired what communications guys called "star topography", which is a fancy phrase for "outlets are usually spidered out of a ceiling light fixture".
The only thing with ungrounded systems is that surge protectors do not work as there is no ground to which the surge would get directed.
Modern electronics generally work fine with ungrounded systems; the grounding wire is not used under normal operation. The only time it may be an issue is with generators and inverters as many of those do not have a true sine wave output
mbrooke
Batteries Included
- Location
- United States
- Occupation
- Technician
I politely disagree. I argue, and I could be wrong, that this is meant to cover new installations only. I can not see the NEC allowing GFCI receptacles to substitute as an EGC only to disallow almost everything with a ground pin. What would make a refrigerator anymore dangerous then say a garage door opener plugged into a none grounded GFCI?
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
151209-0842 EST
JFletcher:
Whether a surge limiter is of no value in a circuit where there is no ECG is dependent upon what is in the surge limiter. Note: there are very few devices that can be real preventers of surge damage. Rather it is better to call them surge or transient limiters.
Transient voltage limiting devices can tolerate large currents only for short time periods. This is a power or energy dissipation problem. Transient limiters, such as MOVs, do not clamp voltage very sharply or very close to the peak nominal AC voltage.
If a 120 V single phase surge limiter contains 3 MOVs with one each connected between hot and neutral, hot and EGC, and neutral and EGC, then if the neutral is intact back to the main panel and earth, then limiting action relative to neutral and earth of a surge on the hot line still exists thru the neutral. If neutral is open, then there is no limiting, but with no EGC does it matter if the load is essentially insulated from earth (a wood floor for example).
I can float a battery powered radio on 10,000 V relative to earth and it will work just fine. This assumes that I do not have some conductive path from the radio to earth.
.
JFletcher:
Whether a surge limiter is of no value in a circuit where there is no ECG is dependent upon what is in the surge limiter. Note: there are very few devices that can be real preventers of surge damage. Rather it is better to call them surge or transient limiters.
Transient voltage limiting devices can tolerate large currents only for short time periods. This is a power or energy dissipation problem. Transient limiters, such as MOVs, do not clamp voltage very sharply or very close to the peak nominal AC voltage.
If a 120 V single phase surge limiter contains 3 MOVs with one each connected between hot and neutral, hot and EGC, and neutral and EGC, then if the neutral is intact back to the main panel and earth, then limiting action relative to neutral and earth of a surge on the hot line still exists thru the neutral. If neutral is open, then there is no limiting, but with no EGC does it matter if the load is essentially insulated from earth (a wood floor for example).
I can float a battery powered radio on 10,000 V relative to earth and it will work just fine. This assumes that I do not have some conductive path from the radio to earth.
.
- Status