# Mailinglist Archive: opensuse (3666 mails)

 < Previous Next >
##### Re: [SLE] PC Crash: Hard Disk Problems
• From: James Knott <james.knott@xxxxxxxxxx>
• Date: Mon, 21 Mar 2005 19:24:09 -0500
• Message-id: <423F65A9.5010805@xxxxxxxxxx>
Randall R Schulz wrote:
James,

On Monday 21 March 2005 03:55, James Knott wrote:

John Sowden wrote:

Note: A ground fault interupter is designed to sense a small
current flow from the hot side (not neutral) to ground. There
should never be a current flow through this path . . .

I thought they detected a current imbalance in the hot & neutral
conductors. It is possible to have leakage to other than ground.

Correct (well, half correct). They're often used when there is no ground reference lead in the building's wiring system. However, there is no other leakage path than to ground (however indirect or--no pun intended--circuitous that path is).

I don't know about where you are, but in North America, a split phase 240/120V system is used, where the center tap of the line transformer is grounded and becomes the neutral. The two "hot" lines are 180 degrees apart, and have 240V between them. It is possible to have leakage to the opposite phase.

GFCIs essentially _infer_ that current is traveling to ground because there's unequal current flow through the two power wires. (Kirchoff's laws to the rescue). The _assumption_ (a conservative one) is that the current is flowing through a human. Humans are notoriously ill-suited for use as electrical conductors.

Quite familiar with Kirchoff. I aced an electrical engineering exam question once, when I had the sense to apply Kirchoff to the problem. I was apparently the only one in the class to do so.

Hmmm... I wonder, how do GFCIs deal with reactive loads? Do they time average the current measurement over several cycles?

The currents should always be exactly opposite, no matter what the reactance is. Remember, you're comparing current, not phase relationship. Also, current in a series circuit with have the same phase relationship at any point.

 < Previous Next >