4 Feb
2011
4 Feb
'11
2:37 p.m.
On Friday 04 February 2011, Tony Duell wrote:
I'd agree. The problem is convincing people to spend the time
tracking down an intermittent fault somewhere within more than a
dozen possible pieces of equipment. Sometimes it's easier to just
wait for it to go BANG!, and pick up the pieces. There's nothing
irreplacable or unique
I see.. There's never time or money to do it right. nut there is time
and moneyu to bodge it when things finally fail catestrophically.
And youwonder wheat's wrong with the world...
"Do it right" is subjective. You may easily spend hundreds of man hours
hunting down a problem that won't show up until there's a major fault,
which costs maybe a few man-hours to fix, and is simple to find once it
has gone "boom".
Basically, an equipment fault which causes current to flow to ground,
but which isn't enough to trip a breaker or blow a fuse.
It's actually required by the US National Electrical Code to have such a
ground-fault interrupter on any power distribution system that exceeds
800A.
I believe that part of the reasoning is that you can have a "dead short"
that's through a thin enough conductor (or strand of multi-stranded
conductor), that you won't draw enough current to blow a fuse (possibly
hundreds of amps), but which you don't want to trip from 90deg out of
phase loads from capacitors or inductors that are acting as power
filters.
Yes, I do understand the difference. My point is that a fault
that's 200A on a supply that can probably supply a fault current
nearing 100kA, isn't that big compared to what's possible, (eg, not
a "dead short").
Ture enough. But what fault would trip this breaker nad not a
current-operated one? What is it protecting against?
Say you have a 18AWG or so "strand" of a thick stranded copper wire, on
a circuit that's fused for 400A or more. The strand (shorted to ground)
will quickly heat up to dangerous levels, and probably won't cause a
high enough current draw to blow the fuse.
Wire insulation that has been knicked by a metal burr in conduit can
cause a similar situation.
Yes, but breakers tend to emulate "slow-blo" fuses more than
"fast-blo"
ones.
quickly...
And, it's most likely a spike, not a continuous
leakage current, eg, a short-lived arc to ground from some worn
out insulating parts.
That is precisely the sort of thing I would repair and not just
turn up the breakers to get round it! Worn-out isulation
(whatever that may mean) is not something I am going to trust. that it could
destroy, and the equipment is generally designed to
contain such a problem to avoid hurting a person who's nearby when
it happens.
Perhaps you can explain to me the point of an earth-leakage breaker
under thsoe conditions... It's nvercurent device (that is, a derive
that cuts th emains if the current in the power-carying conductors
exceeds a certain value). The normal reason for having them over
here is to protect people in the event of an insulation breakdown,
the unbalanced current flowing throughj the person to ground will
trip the breaker.
I doubt that 200A would flow though a person touvhing a live
conductor (if it would,then there'd be no point in the RCD in most
isntalations, the nromal current-operated breakers would trip). And
if it fif, that person would not notice it. Ever.
So I am seriosu. What fault is the 200A earth-leakage breaker
protecting against? In any case a
200A short-term fault isn't all that huge.
Ratings for typical US house circuit breakers are around
10,000A interrupting capability.
That is surely the maximum fault current they will safely break.
Not the current they carry before they trip. Over here most
decent domestic MCBs will break 16kA safely -- that is the peak
current that might flow if there's a dead short across the mains.
But said breakers will trip on a current of 32A, say. And for
whatever it's worth, unless it's a special purpose breaker,
a circuit breaker doesn't trip as soon as the current crosses its
rated current, where "soon" can be as long as minutes depending on
the breaker if you're only overloading by 10% or so. I was amazed
at how long (10-15 minutes) a particular 20A breaker sustained a
25A load before it tripped. Then again, the thermal response of a
breaker tends to match the effects of wires (that it's protecting)
to keep them from overheating.
The same is true of fuses, of course. A 20A fuse does not melt the
instant the currnet rises over 20A. The currnet agianst time (I-t) characteristics of
fuses and circuit
breakers are specified i nthe data sheets (and at least over here
there are British Standards giving hte ones ocmmonly used). It's
often important to allow a higher-than-normal inrush current. I am
sure we've all used anti-surge fuses, for example.
However, I didn;t think that earth-leakage brekaers (RCDs, etc) had
much of a time delay as standard. The idea being that there really
shouldn't be an earth leakage current, not even fopr a faraction of
a second.
Yes, high-current GFIs work the same way.
In addition, high-current breakers usually have a magnetic trip
(possibly with a settable trip point somewhere around 2-4x the rating of
the breaker) that will trip with just a cycle or two of abnormally high
current (eg, short circuit) in addition to the thermal breaker part,
which trips on circuit overload.
--
Purdue University Research Computing --- http://www.rcac.purdue.edu/
The Computer Refuge --- http://computer-refuge.org