[TowerTalk] GFI Circuits

[Jim Lux]

At 03:12 PM 1/21/2005, ny6dx at netscape.net wrote:
>Bud,
>    I think there is more here than meets the eye.A gfi breaker is not rf 
> sensitive,its meerly a current transformer that monitors the current 
> relation between the 120volt lead and the return on the nuetral leg.Only 
> an inbalance in this flow would cause such a failure.
There's a fairly sensitive amplifier to measure the differential current 
(using a transformer with both Line and Neutral running through the core, 
along with a sense winding).  That amplifier or the detection circuitry, 
might be RF sensitive.

In fact, the Leviton GFCI brochure 
(http://www.leviton.com/pdfs/gfci/GFCI_Brochure.pdf) mentions "increased 
noise immmunity: Higher resistance to electrical noise (EMI/RFI) to help 
reduce nuisance tripping" as required by UL starting 1 Jan 2003.

That brochure has some nifty looking devices that might be useful.. GFCIs 
with pilot lights, for instance.

There's also some description of why GFCIs for cord sets (the field day 
sort of application) are different than for normal receptacles.  It has to 
do with a live hot leg with an open neutral on the line side. The tripping 
mechanism won't have any power to operate.



Here's a typical chip used in these things:
http://www.fairchildsemi.com/ds/RV%2FRV4141A.pdf

http://www.fairchildsemi.com/ds/KA%2FKA2807.pdf

The latter requires all of 5 microamps to fire...  It's easy to see how RFI 
could get in there and trip it.  Don't forget that these are a cost 
sensitive device: the mfr's not going to put any parts or filtering in 
there unless absolutely needed (viz. the recent UL rule change).