Any ideas on this. Two year old house. Kitchen GFCI tripped normally with my tester unless I was plugged into the actual GFCI. Then, no trip. The built in test button on the GFCI worked fine. Never seen this before. What would cause it?
Not grounded.
Bingo Bob…I see it all the time.
That’s why you always test the actual GFCI outlet. Sometimes they show up grounded, but still don’t trip. Who knows why. I write it up stating that it would not trip using a dedicated GFCI test device.
Does this mean that it would trip when you tested downstream devices but not the actual GFCI receptacle itself? If so, I agree with the others, the GFCI receptacle itself is un-grounded.
There is no reason for a 2 year old house to have an ungrounded receptacle.
I know Joe said he didn’t know why, but does anyone else know why that doesn’t show up as ungrounded on the tester?
It can be set up as a false ground .
The others may actually be grounded.
False ground at outlet.
A GFCI monitors the amount of current flowing from hot to neutral. If there is any imbalance, it trips the circuit. In some areas, upgrading 2 prong outlets (in older homes with a 2 wire electrical system) with 3 prong GFCI outlets is acceptable. Without a ground, a GFCI should still trip. Something else is causing it not to trip.
Some GFCI testers can simulate too weak an electrical “leak” or imbalance to trip some GFCI’s.
Did it trip when you mashed the test button?
False ground.
I cant believe I am about to say this but, I…I…I…agreewithBob.:mrgreen:
Yes…Yess…Yessssssss
Join me.ha…haaa.haaaaaaa!
Sorry guys, I happen to disagree. The GFCI does not use the equipment grounding conductor in the sensing circuit. A GFCI will protect non-grounded circuits and according to the NEC Sec. 210-7(d), GFCI can replace two-wire (ungrounded) receptacles.
Being ungrounded should not cause a GFCI NOT to trip. However, I am still not sure why the downstream outlets will trip but not the first, and I don’t see how being ungrounded or having a false ground would cause this either. Please explain!
Sloppy work would do it.
Bob, do you think it might not be grounded?! 
Could be a false ground .
Ground to neutral.
Just a theory.
Button test is the only one that counts for failure of the devise from everything I have read so far…
You’re correct the GFCI device does not use the EGC for a self test with the TEST button. However if you inserted a GFCI tester into the GFCI receptacle and the receptacle were ungrounded the tester would not trip the GFCI receptacle. If you plugged your same tester into a downstream receptacle (one fed as a feed through from the GFCI) that was grounded and pressed the TEST button on the tester it would trip the upstream GFCI receptacle. In any case, grounded or ungrounded, if the GFCI TEST button is pressed it should trip the receptacle if it’s operating normally.
I had one recently where the lights on my tester indicated correct wiring, but the button would not trip it. The GFCI’s test buttons did, but not mine. The tester worked fine before, and after, just not for this GFCI.
Ok, I just had to see for myself. So, I pulled the cover from the GFCI outlet in my kitchen, took the ground wire off from the ground terminal and tested the GFCI with my tester. Sure enough it would not trip, but the others downstream would. I apologize for disagreeing with you guys earlier, you were correct. Had I installed a jumper wire, it would have displayed correct wiring on the tester, but of course would not actually be grounded. A false ground is absolutely the correct answer to the first question of this post.
Thanks Mr. Meier for the detailed explanation above.