On Sun, 30 Nov 2008 19:11:56 +0000, Tom Anderson
wrote:

On Sun, 30 Nov 2008, David Hansen wrote:

On Sun, 30 Nov 2008 16:50:43 +0000 someone who may be Steve
Fitzgerald ] wrote this:-

http://www.tubelines.com/whatwedo/tr...guidance/LU_sh
ort_circuiting_device.pdf

Which sadly in this case, the SCD was the cause of the problem.

I guessed that it was one of a series of problems which caused the
voltage from a still energised positive rail to be transferred to the
negative rail. Presumably there was some sort of broken connection or
open circuit breaker/fuse in the feed to the negative rail.

Ah, so although there was no potential between the two power rails,
there was a potential between them and the ground?

Really, you need to connect the two rails to each other and also to the
ground. I was thinking the running rails would make a good ground
substitute here.

Only on one running rail of a 3-rail system where there is intentional
permanent bonding suitable for traction currents (and even then it
isn't necessarily at the same PD as the local earth). On a LU 4-rail
system the running rails are not "earthed" WRT traction supplies and
have no certain/permanent low-resistance path to earth even if they
did when a 3-rail SCD (positive-to-running rail) was first applied.
Think of the earth wire in a domestic system - this normally carries
no current but still has a minimum current rating and maximum
permitted impedance to the supply earth as it is an intentional path,
albeit with a limited expectation of actual use.