myredracer wrote:
The link to a page from a hot tub supplier isn't exactly something I'd call credible.
OK, I'll stay out of tech stuff. The linked GFCI article on the hot tub supplier's site was originally on codecheck.com. It was removed some time ago from codecheck.com but survives on other sites. The author is a prof at U of Pennsylvania, has published hundreds of tech articles, and is a principal contributor at sci.electronics.repair and repairfaq.org. His GFCI article is referenced often on the Mike Holt forum, and I have not found any rebuttal to it there. For me, I consider it credible.
Breaking the "no tech stuff" on edit, there's a newer GFCI chip from Fairchild that doesn't use the 120 Hz. signal. It still uses a 2nd coil which is normally not energized. When there's a downstream ground-neutral fault, the interaction of the 2nd coil with the main sense coil results in the sense amplifier oscillating, indicating the ground-neutral fault. I have a fairly new Leviton GFCI that uses a Fairchild chip. Older Leviton GFCIs used the LM1851 chip that uses the 120 Hz. signal. I don't know when the chip change occurred or if anybody uses the LM1851 any more. Here's a "How GFCIs work" powerpoint presentation that describes the 120 Hz signal technique:
McGraw Hill GFCI presentationSlides 9 & 10 cover the grounded neutral detection.