I just replaced a regular outlet by a GFCI outlet. A GB outlet tester
with GFCI tester shows that there is no ground (just the same as before
I changed the outlet -- I'll try to solve that problem later: there is a
bare copper wire connected to the metal box and to the outlet's green
screw, but it must be disconnected at the other end), BUT pressing the
outlet tester's GFCI Test button still trips the breaker. I thought that
a ground connection was necessary for the test button to work.
According to the diagrams here
the test button diverts some current from hot to ground to cause the
hot/neutral conductors to not carry the same amount of current. Without
a ground, how would the test button work? Where does the diverted
Doh...should learn to research first. Found an alternate design where
the test button diverts some current around the sensing toroid back to
neutral such that the toroid sees unbalanced current even though it is
actually balanced for the receptacle as a whole.
Works just fine with no real ground.
That scheme only applies to the built in test circuit in the GFCI
itself. A plug in tester does indeed need a ground in order to work but
the ground only has to be good enough to carry six milliamperes. That
would not make it a satisfactory low impedance pathway back to the
source via the main bonding jumper.
"This alternating current stuff is just a fad. It is much too dangerous
Something that this whole discussion has prompted me to wonder...has
anyone ever heard of a tester (with a long heavy cable) that would plug
into the service panel directly that you could use to *really* test
outlets? Could be used to test non-grounded GFCIs, test for floating
ground, put 13A through each of the leads, measure voltage drop under
load, etc...? Seems like it might be useful to highlight possibly
So when you push the test button on the outlet tester it trips? (or the
built-in GFCI test button?)
The external tester leaves out bypassing the toroid, I assume the GB
tester shorts hot-to-gnd through a resistor, so there must be enough of
a leakage current through the 'open' ground to trip the GFCI (as far as
I can reckon)
See my other post. If the self-test shunts some current from hot to
neutral bypassing the toroid, you could have a self-test that works fine
while the external tester wouldn't be able to work due to a lack of a
And the diagram appears to show this incorrectly. It shows both sides
of the resistor connecting to the LINE side of the transformer, when
you it needs to connect to one of each side. The real GFCI I looked at
HomeOwnersHub.com is a website for homeowners and building and maintenance pros. It is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.