Maker Pro
Maker Pro

GFCI Failures + Gadgets

J

Jim Thompson

Jan 1, 1970
0
Went outside to add another electrical circuit at the BBQ and
discovered that all of my nice wire-tagging had faded to non-readable.

So I couldn't figure out which common went with which hot.

Then it dawned on me to simply trip the GFCIs which releases *both*
hots and commons so a simple ohm-meter check would do the trick.

NOT! Both GFCIs wouldn't trip when I pressed the test buttons :-(

Went to Radio Shack and bought a Receptacle/GFCI tester for $5.99.

At the same time I noticed a gadget to locate breakers... bought it
also ($29.95).

Went home and tested the GFCIs... both have failed :-( Tester was
verified on some indoor GFCIs... it would trip them.

Both bad units are outdoors, so maybe it was the heat (they both face
the western sun... it gets over 120°F here :). Any other ideas about
why they fail?

The breaker locator is neat... plug a sender unit into an outlet, then
scan the breakers... works like a champ.

...Jim Thompson
 
W

w_tom

Jan 1, 1970
0
Demonstrated is but another reason why homes require a
'whole house' surge protector on each incoming utility wire.
GFCIs are electronics that contain effective protection.
Protection that can be overwhelmed if incoming transients (ie
lightning) are not earthed before entering the building. IOW
GFCI failure is a symptom that a whole household of appliances
have ineffective protection. Maybe the home does not have the
proper central earth ground as well as no 'whole house'
protector. But GFCI failures can be created by incoming,
common mode transients. I have been told this type of failure
damages as many as 50% of all GFCIs in central FL.

$5+ for a GFCI tester? Its nothing more than a switch and
maybe a 15K resistor that connects hot (black) wire to safety
ground (green) wire. That same tester is what the TEST button
does - already inside the GFCI.
 
J

Jim Thompson

Jan 1, 1970
0
Demonstrated is but another reason why homes require a
'whole house' surge protector on each incoming utility wire.
GFCIs are electronics that contain effective protection.
Protection that can be overwhelmed if incoming transients (ie
lightning) are not earthed before entering the building. IOW
GFCI failure is a symptom that a whole household of appliances
have ineffective protection. Maybe the home does not have the
proper central earth ground as well as no 'whole house'
protector. But GFCI failures can be created by incoming,
common mode transients. I have been told this type of failure
damages as many as 50% of all GFCIs in central FL.

There *is* a whole-house surge-protector. All the GFCIs inside check
out OK, so I'm guessing it's a heat issue only.
$5+ for a GFCI tester? Its nothing more than a switch and
maybe a 15K resistor that connects hot (black) wire to safety
ground (green) wire. That same tester is what the TEST button
does - already inside the GFCI.

I earn $5.95 in less than 2 minutes designing ICs. I don't build
*anything* anymore except for fun or if the function is unavailable...
plus this tester also verifies phasing and no lost connections ;-)

...Jim Thompson
 
L

Louis Bybee

Jan 1, 1970
0
w_tom said:
Demonstrated is but another reason why homes require a
'whole house' surge protector on each incoming utility wire.
GFCIs are electronics that contain effective protection.
Protection that can be overwhelmed if incoming transients (ie
lightning) are not earthed before entering the building. IOW
GFCI failure is a symptom that a whole household of appliances
have ineffective protection. Maybe the home does not have the
proper central earth ground as well as no 'whole house'
protector. But GFCI failures can be created by incoming,
common mode transients. I have been told this type of failure
damages as many as 50% of all GFCIs in central FL.

$5+ for a GFCI tester? Its nothing more than a switch and
maybe a 15K resistor that connects hot (black) wire to safety
ground (green) wire. That same tester is what the TEST button
does - already inside the GFCI.



Not exactly the same. As I understand it the internal GFCI test switch will
work without a ground conductor connected because the connection is effected
between the hot conductor after the GFCI Device, and the Neutral before the
GFCI. The aftermarket testors wont trip the GFCI Device without a ground
conductor connected to it.

There are also usually indicators that allow some level of troubleshooting
as to the proper wiring of a receptacle.

Louis--
*********************************************
Remove the two fish in address to respond
 
W

w_tom

Jan 1, 1970
0
Just because some GFCIs are working does not mean a surge
could not exist. These GFCIs are outside. Using buried
wire? IOW if the exterior (exterior) GFCI made a better earth
ground than the central earth ground for 'whole house'
protector, then surge would also use GFCI as a destructive
path to earth. If mounted on side of building, then this
scenario is not so likely.

Operating temperature for older versions GFCI chips is
typically up to 70 degrees C. Chip consumes near zero watts.
IOW heat is not a likely reason for failure for those type.
Have not seen datasheets for newer type GFCIs. But then is
that receptacle getting so warm that IC would exceed 70 degree
C?

How good (and short) that that ground for the 'whole house'
protector. Is there any good earthing path via those GFCIs?
Perhaps are those GFCIs located on the far side of the
building?
 
G

Gary Tait

Jan 1, 1970
0
Not exactly the same. As I understand it the internal GFCI test switch will
work without a ground conductor connected because the connection is effected
between the hot conductor after the GFCI Device, and the Neutral before the
GFCI. The aftermarket testors wont trip the GFCI Device without a ground
conductor connected to it.


It will, the test button on a GFCI usually, through a resistor,
connects the GFCI protected hot to the unprotected neutral.
 
B

BinaryBillTheSailor@Sea++.com

Jan 1, 1970
0
Just because some GFCIs are working does not mean a surge
could not exist. These GFCIs are outside. Using buried
wire? IOW if the exterior (exterior) GFCI made a better earth
ground than the central earth ground for 'whole house'
protector, then surge would also use GFCI as a destructive
path to earth. If mounted on side of building, then this
scenario is not so likely.

Operating temperature for older versions GFCI chips is
typically up to 70 degrees C. Chip consumes near zero watts.
IOW heat is not a likely reason for failure for those type.
Have not seen datasheets for newer type GFCIs. But then is
that receptacle getting so warm that IC would exceed 70 degree
C?

In direct sun during hot weather? That wouldn't be surprising at all.

BB
 
N

N. Thornton

Jan 1, 1970
0
w_tom said:
Demonstrated is but another reason why homes require a
'whole house' surge protector on each incoming utility wire.
GFCIs are electronics that contain effective protection.
Protection that can be overwhelmed if incoming transients (ie
lightning) are not earthed before entering the building. IOW
GFCI failure is a symptom that a whole household of appliances
have ineffective protection. Maybe the home does not have the
proper central earth ground as well as no 'whole house'
protector. But GFCI failures can be created by incoming,
common mode transients. I have been told this type of failure
damages as many as 50% of all GFCIs in central FL.

Little surge protectors offer very little protection against
lightning. That is a common misconception.

$5+ for a GFCI tester? Its nothing more than a switch and
maybe a 15K resistor that connects hot (black) wire to safety
ground (green) wire. That same tester is what the TEST button
does - already inside the GFCI.

indeed.


Regards, NT
 
J

James Meyer

Jan 1, 1970
0
Both bad units are outdoors, so maybe it was the heat (they both face
the western sun... it gets over 120°F here :). Any other ideas about
why they fail?
Are there both indoor and outdoor grade GFCI breakers available like
there are for ordinary receptacles?

Jim
 
J

Jeff

Jan 1, 1970
0
Jim Thompson said:
Went outside to add another electrical circuit at the BBQ and
discovered that all of my nice wire-tagging had faded to non-readable.

So I couldn't figure out which common went with which hot.

Then it dawned on me to simply trip the GFCIs which releases *both*
hots and commons so a simple ohm-meter check would do the trick.

NOT! Both GFCIs wouldn't trip when I pressed the test buttons :-(

Went to Radio Shack and bought a Receptacle/GFCI tester for $5.99.

At the same time I noticed a gadget to locate breakers... bought it
also ($29.95).

Went home and tested the GFCIs... both have failed :-( Tester was
verified on some indoor GFCIs... it would trip them.

Both bad units are outdoors, so maybe it was the heat (they both face
the western sun... it gets over 120°F here :). Any other ideas about
why they fail?

Rip one apart and see what failed. They usually have screws on the back. All
they are is a current transformer with both the hot and the neutral going
through it. If the current in the neutral and the hot is not exactly the
same, a current is produced in the current transformer. Most GFI's trip at a
5 mA differential, meaning a 5 or more mA current leakage from the neutral
or hot will trip the GFI. Perhaps the solder joints failed from the excess
thermal cycling?
 
L

Louis Bybee

Jan 1, 1970
0
Gary Tait said:
It will, the test button on a GFCI usually, through a resistor,
connects the GFCI protected hot to the unprotected neutral.


If you read my answer above, this is exactly what I indicated. In the second
part of my reply I suggested that an aftermarket tester (not the one on the
GFCI) wouldn't trip a GFCI where there wasn't a ground conductor connected
to it (try it).

Louis--
*********************************************
Remove the two fish in address to respond


 
G

Gary Tait

Jan 1, 1970
0
Are there both indoor and outdoor grade GFCI breakers available like
there are for ordinary receptacles?

Jim

If you are asking about a breaker that fits in your panel, just about
all current styles have a GFCI breaker.
 
J

James Meyer

Jan 1, 1970
0
If you are asking about a breaker that fits in your panel, just about
all current styles have a GFCI breaker.

No. I was asking about the outlet receptacle type of GFCI.

But adding a single ground fault breaker per circuit to the panel would
seem to be a much better solution than adding multiple outlet type ground fault
breakers.

Jim
 
F

Fred Abse

Jan 1, 1970
0
It will, the test button on a GFCI usually, through a resistor, connects
the GFCI protected hot to the unprotected neutral.

That's just the same as loading the circuit, it won't work. GFCIs monitor
current in *both* hot and neutral, which should be the same.

You need to create an imbalance current by introducing a resistance
between hot and *ground* 5mA might do it. 30mA _must_ do it, or the
protection is no damn good.
 
J

JeffM

Jan 1, 1970
0
Most GFI's trip at a 5 mA

That sounded too high, so I googled for confirmation. You're right.

When I worked in the hospital,
I remember the Electrical Safety Officer coming around with his cart
and checking every 110Vac gizmo. His spec was 15uA.
(You would marvel at this collection of cut-off molded 110V plugs.)
 
J

Jeff

Jan 1, 1970
0
JeffM said:
That sounded too high, so I googled for confirmation. You're right.

When I worked in the hospital,
I remember the Electrical Safety Officer coming around with his cart
and checking every 110Vac gizmo. His spec was 15uA.

Whoa! 15uA - Some insulation will collectively leak that at 120V AC
especially on a humid day, and then there's stray capacitance leaking to
ground!
 
J

Jeff

Jan 1, 1970
0
James Meyer said:
No. I was asking about the outlet receptacle type of GFCI.

But adding a single ground fault breaker per circuit to the panel would
seem to be a much better solution than adding multiple outlet type ground fault
breakers.

Breakers are the best, however they generally cost 4 to 8 times as much, and
have the inconvenience of needing to go to the panel every time they trip
(and they can trip often when using power outside). The GFI plugs also have
terminals to wire additional plugs, lights, etc to protect things later on
in the circuit.
 
M

Mark Jones

Jan 1, 1970
0
In news:[email protected] (Jeff):
Whoa! 15uA - Some insulation will collectively leak that at 120V AC
especially on a humid day, and then there's stray capacitance leaking to
ground!

Yeah but, in the desert southwest, there ain't much in the way of stray
humidity...
 
James said:
No. I was asking about the outlet receptacle type of GFCI.

But adding a single ground fault breaker per circuit to the panel would
seem to be a much better solution than adding multiple outlet type ground fault
breakers.

A GFCI receptacle protects both itself and the circuit
connected to the load side of the receptacle.
 
Top