Never seen this one before. GFCI receptacle test and reset buttons work normally. However, when a tester is used, the breaker trips instead of the receptacle. The same thing happens on other outlets protected by the GFCI receptacle. Any ideas why?
The common assumptions is that all GFCIs are receptacles. If an outlet needs to be GFCI protected, and you don't see a funny receptacle with a Test and Reset button, therefore somebody messed up.
Actually that is not true. GFCI is actually a "zone of protection" that can be projected to any section of a circuit. And GFCI devices which do that come in a large variety of packages:
- GFCI + receptacle of course
- Standalone GFCI (deadfront)
- GFCI that is a switch (switch-rated deadfront)
- GFCI + receptacle combined with a switch
- GFCI circuit breaker
- GFCI + AFCI + circuit breaker
In your case, my guess is your breaker is a GFCI breaker. That makes a certain amount of sense, since I gather that receptacle is in a location that needs GFCI protection.
If I am correct, and a "Test" button and markings on the breaker will confirm it, then the GFCI receptacle is entirely redundant and can be replaced with a plain receptacle.
I can't call the person a fool for installing it, because there was in fact a GFCI code violation here. Anytime a receptacle is protected from another location, it must be labeled "GFCI Protected" in some manner not handwritten. See NEC 110.3(B) and instructions 8(C). See also 406.4(D)(2).
Why does the GFCI receptacle's test not trip the GFCI breaker? Because the GFCI receptacle TEST button does not create a genuine ground fault, because it can't count on ground being present. So the GFCI creates a "pseudo" ground fault between "protected zone" hot and supply-side neutral, inducing a differential current. This looks like normal hot-neutral load to the upstream GFCI.