Joel Corwith said:
I've run into units in which the test switch trips the unit, but a GFCI
tester doesn't. I don't trust the test buttons anymore.
Joel. phx
Had you considered the possibility that your GFCI Tester was defective, and
the test button on the GFCI was functioning properly?
If you have the skills to work with live circuits safely (there is the
potential for a lethal shock) you could construct a circuit that could tell
you if the GFCI was functioning as intended.
With all live components insulated, and enclosed in a box you could assemble
a circuit where a three conductor cord with the appropriate male plug on the
other end was brought into the box with a 6.8k resistor connected to the hot
wire to limit current to approximately 18 ma. The other lead of the resistor
would be connected to a 100k pot. From the wiper of the pot a connection
would be made to a shrouded banana receptacle. A second shrouded banana
receptacle would connect to the ground conductor. With a multi meter
connected to the banana receptacles, and set to the current mode you could
read the current flowing in the circuit when plugged into the GFCI device.
With the pot adjusted to one extreme approximately 1 ma would flow. At the
opposite extreme approximately 18 ma would flow. With nothing else connected
to the GFCI except your test circuit (either the other receptacle on the
GFCI, or to the load terminals) you could slowly increase the current
flowing to observe where the device tripped. Recording that reading, and
reconnecting the other loads to the GFCI would allow you to repeat the test
and subtract to determine the approximate leakage of the load normally
connected to the GFCI device. I believe a GFCI device designed for
convenience outlets is specified to trip between 4 & 6 ma.
Measuring the current load your tester places on the GFCI circuit could also
indicate it's place in the puzzle.
Louis