According to SpamFree :
In none of the above do you make reference to anything we can check or
to anything which explains the apparent anomaly of a perfectly good
steel strip connecting the source of the fault with the building
ground not working properly if wound in the form of a spiral but being
OK if a ridiculously small gauge wire follows along the spiral. The
proffered explanation of inductive resistance seems to be definitively
crushed by Chris Lewis who in the earlier discussions was supported by
other posters who had presumably done the same calculations. In any
event no one pointed to errors in his reasoning.

Let's not start all this again, and look at something else - Canadian CEC
hasn't permitted cable sheath to be a ground for at least 30 years,
has _never_ approved armored cable with that "ridiculously small gauge
[bonding] wire", and has insisted for at least 30 years on a full size
copper ground conductor.

Cable armor is physical protection and little more.

The facts are simple: the cross-sectional area of cable armor is small,
perhaps not even as much as a copper wire. It's steel, not copper. Steel
rusts. Even when galvanized, the edges rust. Even if aluminum, the
manufacturing process will not produce long-term high conductance joints.

It's small cross-sectional area to begin with. Rust makes it smaller. It's
really long. It's brittle. An impact can destroy the electrical conductivity
if it cracks the strip. One must NOT rely on the conductivity of cable
sheath or box clamps for ground continuity.

There's a reason that current carrying conductors are copper, not steel.

IIRC, US "MC" cable is armored with a full size copper ground. Use _that_.

In Canada, that's all we're allowed to use.

Tom is right. Cable armor is a lousy ground. It's just that reactive impedance
is _not_ the reason why it's lousy.

[If I recall my calculations right, several hundred feed of cable armor will
have an equivalent impedance of a few microhenries. At 60hz, the reactive
impedance is insignificant - nowhere near enough to affect breaker trip.
It's like saying "skin effect" matters at 60hz. Skin effect exists. But
at 60hz it can be totally ignored.]
--
Chris Lewis, Una confibula non set est
It's not just anyone who gets a Starship Cruiser class named after them.